Metadvice Medical Mysteries: Covid 19 - Part II
This Metadvice blog series explores some of the mysteries surrounding the complexities of SARS-CoV-2 infection on people and healthcare systems, and how advanced analytics and technologies like artificial intelligence (AI) can help navigate these complexities.
The novel coronavirus known as severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) invades the human body by docking with angiotensin-converting enzyme 2 (ACE2). This protein is widely expressed in human tissues. Whether higher expression of ACE2, determined indirectly via levels of the circulating soluble form of the receptor, accounts for worse outcomes in men relative to women requires further evaluation. Once infected with SARS-CoV-2, affected individuals may exhibit a variety of symptoms. The majority (~80%) display only mild symptoms (fever, cough, sore throat, headache, fatigue) and recover without the need for hospitalization. The remaining ~20% become seriously ill, according to the WHO. In severe cases, respiratory complications may include pneumonia, sepsis, respiratory failure and acute respiratory distress syndrome (ARDS). In some patients, the immune response is manifested as a cytokine storm.
While the activation of coagulation during inflammation and the interaction between inflammation and coagulation are not completely understood it is apparent that these responses can lead to adverse clinical outcomes via venous, arterial and capillary thromboses. Studies to date have shown that blood clots appear in ~20-30% of critically ill COVID-19 patients. COVID-19-associated coagulopathy is manifested primarily as organ dysfunction, haemorrhagic events being less frequent. Laboratory markers of extensive fibrin formation include increased circulating levels of D-dimers and fibrin/fibrinogen degradation products which have been shown to predict mortality in hospitalized patients with COVID-19. Extensive pulmonary microthrombosis is implicated in one of the early clinical mysteries to be identified in COVID-19, i.e. the marked yet asymptomatic degrees of hypoxia exhibited by some patients. The term ‘silent hypoxia’ has been applied to this situation where pulse oximetry reveals surprisingly severe hypoxaemia in patients who do not appear to be unduly symptomatic. Extrapulmonary manifestations of COVID-19 include thrombotic complications, cardiovascular events, acute kidney injury, gastrointestinal symptoms, hepatocellular injury, hyperglycaemia and ketosis, neurological pathology, and dermatological complications.
Beyond pre-existing respiratory disorders and conditions associated with impaired immune responses, certain comorbidities are known to affect the severity and progression of COVID-19. Here we consider disorders that have been of particular early interest to Metadvice.
Predictably, cancer patients are at higher risk of severe COVID-19 illness and mortality compared to the general population. These adverse outcomes are attributed to the debilitating effects of cancer per se and may be exacerbated by the effect of cancer treatment on the integrity of the immune system.
Patients with haematological and lung cancers appear to be at particular risk. However, data from case studies suggest that COVID-19-related mortality in cancer patients appears to be principally driven by age, gender, and comorbidities.
Prominent among the conditions that were identified in early reports emanating from Wuhan included hypertension, diabetes and cardiovascular disease. These disorders, which are closely interrelated, are strongly age-related and serve to impair resilience to intercurrent illnesses in affected individuals. Accordingly, it is hypothesised that the high prevalence of these comorbidities may help explain why the elderly have the highest mortality rates from COVID-19.
Obesity is a common factor underpinning this constellation of cardiometabolic risk factors and was itself identified at an early point as a potent marker for severe COVID-19 infection. The central importance of obesity on outcomes became apparent at an early stage in the COVID-19 pandemic. A dose-dependent effect on likelihood of being admitted to critical care with increasing body mass index (BMI) has been reported. Non-white populations may be especially vulnerable as a consequence of their predisposition to cardiometabolic disease at lower BMI levels. While the mechanisms mediating the elevated risks associated with obesity remain incompletely delineated, reduced cardiorespiratory reserve with immune dysregulation is postulated to mediate the progression to critical illness with organ failure. The pro-thrombotic state induced by obesity may be of relevance to the pathophysiology of COVID-19.
The link between hypertension and COVID-19 has attracted attention in part because of the potential impact of certain classes of antihypertensive medications on susceptibility and risk of an adverse clinical course. The latter include angiotensin converting enzyme (ACE) inhibitors and angiotensin receptor blockers (ARBs). As discussed, the entry route for SARS-CoV-2 is via the ACE2 receptor. Concerns about these widely used medications arose from considerations of biological plausibility allied to the aforementioned overrepresentation of patients with hypertension and other cardiovascular comorbidities among patients with COVID-19 who have poor clinical outcomes.
At present, the consensus view based on the available observational data is that no association exists between long-term use of ACE inhibitors or ARBs and either the severity of COVID-19 or clinical outcomes.
In England, the overall death rate for people with diabetes doubled during the early stages of the pandemic. The risk of COVID-19-related mortality in people with either type 1 and type 2 diabetes was independently associated with levels of hyperglycaemia and obesity. Factors including age, sex, ethnicity, comorbidities (hypertension and cardiovascular disease), obesity, and a pro-inflammatory and pro-coagulative state are considered to contribute to the risk of worse outcomes in patients with diabetes.
The association between hyperglycaemia and poor clinical outcomes has now been extended to include consideration of patients not known to have diabetes prior to developing COVID-19 or who develop acute hyperglycaemia in the setting of severe infection. In a recent report, from Wuhan, nearly half of patients hospitalized with COVID-19 without a prior diagnosis of diabetes had hyperglycemia. Furthermore, an elevated fasting blood glucose on admission was an independent predictor of mortality at 28 days.
Whether COVID-19 causes metabolic disturbances out of proportion to the severity of the acute illness (prolonged ketosis, high insulin requirements reflecting insulin resistance) is presently uncertain. Another intriguing hypothesis is that COVID-19 might induce new-onset diabetes in susceptible patients. Plausible mechanisms have been proposed through which SARS-CoV-2 might disturb glucose metabolism, notably evidence of expression of the ACE2 receptor in metabolically active tissues including the insulin-producing b-cells of the pancreatic islets. Amid the complexities of the bidirectional relationship between COVID-19 and diabetes – which will be a challenge to unravel – the impact of tight glucose control during the acute illness remains unclear; clinical trials to address this important therapeutic question are in progress.
Turning to cardiovascular disease, individuals with cardiac disease (coronary heart disease, heart failure) have a higher risk of developing more severe COVID-19 with a several-fold elevation in mortality. Patients with pre-existing heart disease may sustain a myocardial infarction or develop congestive heart failure precipitated by COVID-19. Elevations of the circulating cardiac biomarker tropopin I may reflect a multiplicity of pathogenic processes. Rapid deteriorations in cardiovascular status are considered to reflect the effect of the severe viral illness on the myocardium compounded by reduced oxygen delivery when myocardial oxygen requirements are increased, and the propensity for blood clot formation.
Acute myocarditis (inflammation of the heart muscle) has been reported. While the incidence of this complication, which may be underdiagnosed in critically ill patients with COVID-19, remains uncertain, two recent studies have provided new insights into cardiac involvement. The first, using magnetic resonance imaging in patients recovering from COVID-19, demonstrated a high prevalence of left ventricular dysfunction and ongoing inflammation; the longer-term clinical implications of these findings remain to be determined. The second study, based on post-mortem examinations of patients who succumbed to pneumonia, showed the presence of viral RNA in the myocardium, raising the possibility of direct myocardial injury.
One of the major needs of healthcare systems and health professionals during the current pandemic and anticipated future waves is the ability to identify and stratify at-risk individuals. Metadvice’s neural network technology has the potential to improve this stratification using deep analysis aimed at identifying novel risk factors. Our technology is also suitable for detailed analysis of the impact of comorbid risk factors, the risk of complications, and the identification of patients for clinical trials.
Since the first reported case of COVID-19 in late December of 2019, a tremendous amount of information about SARS-CoV-2 has accumulated over a short period. The undeniable reality, however, is that knowledge concerning crucial issues such as vulnerability to infection and risk of progression to severe illness remains far from complete. It is clear that translating the complex and often incomplete observational (i.e. real world) data into clinical practice and public health initiatives in a dynamic pandemic will continue to be a major challenge. On a more positive note, encouraging results from therapeutic clinical trials are already being implemented into clinical practice. These advances are considered in another article in this series.