Patients with COVID-19 who have high levels of the steroid hormone cortisol on admission to hospital have a substantially increased risk of dying, U.K. researchers have discovered.
Waljit S. Dhillo, MBBS, PhD, head of the division of diabetes, endocrinology and metabolism at Imperial College London, and colleagues studied 535 patients admitted to major London hospitals. Their article was published online June 18 in Lancet Diabetes & Endocrinology.
“Our analyses show for the first time that patients with COVID-19 mount a marked and appropriate acute cortisol stress response,” said Dr. Dhillo and colleagues.
Moreover, “high cortisol concentrations were associated with increased mortality and a reduced median survival, probably because this is a marker of the severity of illness.”
So measuring cortisol on admission is potentially “another simple marker to use alongside oxygen saturation levels to help us identify which patients need to be admitted immediately, and which may not,” Dr. Dhillo noted in a statement from his institution.
“Having an early indicator of which patients may deteriorate more quickly will help us with providing the best level of care as quickly as possible. In addition, we can also take cortisol levels into account when we are working out how best to treat our patients,” he said.
However, it’s important to note that this means – particularly in the wake of the RECOVERY trial reported last week – that “in the early part of the disease you don’t need steroids,” he said.
In contrast to SARS, no adrenal insufficiency with COVID-19
Cortisol levels when healthy and resting are 100-200 nmol/L and nearly zero when sleeping, the researchers explained.
They decided to examine cortisol levels because, although physiological stress from critical illness normally increases levels of the hormone, the prior coronavirus, severe acute respiratory syndrome coronavirus (SARS-CoV), had the opposite effect and induced cortisol insufficiency in some patients.
“We would have said we’re not quite sure” what effect SARS-CoV-2 is having on cortisol levels, “so that’s why we collected the data,” Dr. Dhillo said in an interview.
The researchers studied patients admitted to three large London teaching hospitals between March 9 and April 22 with a clinical suspicion of SARS-CoV-2 infection. All patients had a standard set of blood tests, including full blood count, creatinine, C-reactive protein, D-dimer, and serum cortisol.
After exclusions, the team assessed 535 patients admitted over the study period who had baseline cortisol measured within 48 hours of admission.
Of these, 403 patients were diagnosed with COVID-19 based on a positive result on real-time polymerase chain reaction testing (88%) or a strong clinical and radiological suspicion, despite a negative test (12%).
In total, 132 (25%) individuals were not diagnosed with COVID-19.
Patients with COVID-19 were a mean age of 66.3 years, and 59.6% were men.
Mean cortisol concentrations in patients with COVID-19 were significantly higher than those not diagnosed with the virus (619 vs 519 nmol/L; P < .0001).
And by May 8, significantly more patients with COVID-19 died than those without (27.8% vs 6.8%; P < .0001).
Doubling of cortisol levels associated with 40% higher mortality
Multivariate analysis taking into account age, presence of comorbidities, and laboratory tests revealed that a doubling of cortisol concentrations among those with COVID-19 was associated with a significant increase in mortality, at a hazard ratio of 1.42 (P = .014).
And patients with COVID-19 whose baseline cortisol level was >744 nmol/L had a median survival of just 15 days, compared with those with a level ≤744 nmol/L, who had a median survival of 36 days (P < .0001).
The team notes that the cortisol stress responses in their patients with COVID-19 ranged up to 3,241 nmol/L, which is “a marked cortisol stress response, perhaps higher than is observed in patients undergoing major surgery.”
Of interest, there was no interaction between cortisol levels and ethnicity in their study; a subsequent analysis of the data stratified by black, Asian, and other minority ethnicities revealed no significant differences.
The team note that their results will need to be reproduced in other populations.
“Any potential role for cortisol measurement at baseline and later during an inpatient stay with COVID-19 as a prognostic biomarker, either by itself or in combination with other biomarkers, will require validation in a prospective study.”
Implications for treatment: Reserve dexamethasone for critically ill
Dr. Dhillo explained that, because their findings indicate that people initially infected with COVID-19 do mount an appropriate stress (cortisol) response, it is important that people properly understand this in the wake of the RECOVERY trial, reported last week.
The trial showed that the widely available steroid dexamethasone significantly reduced mortality among severely ill COVID-19 patients in the intensive care unit when given at a supraphysiologic dose of 6 mg.
But it would be hazardous for anyone to self-medicate with steroids at an early stage of COVID-19 because that would further increase cortisol levels and could suppress the immune system.
“For the average person on the street with COVID-19,” excess steroids will make their symptoms worse, Dr. Dhillo explained, adding this is important to emphasize because dexamethasone, and similar steroids, “are cheap and likely available on the Internet, and so misunderstanding of the RECOVERY trial could have serious implications.”
But once patients are very sick, with “inflammation in their lungs” and are in the intensive care unit, and often on ventilators – which is a very small subgroup of those with COVID-19 – it becomes a very different story, he stressed.
“RECOVERY shows clearly there seems to be a benefit once you need oxygen or are on a ventilator, and that makes sense because [dexamethasone] is going to be an anti-inflammatory,” in this instance when the “lungs are full of water.”
“But in the early days you definitely don’t need it and it could be harmful,” he reiterated.
The study is funded by the U.K. National Institute for Health Research and Medical Research Council. The authors have reported no relevant financial relationships.
This article first appeared on.