This is the authors' abstract. We don't add key points for this paper.
Perhaps one of my earliest memories of seeing what could be the future of medicine is watching a doctor apply a medical tricorder while examining a patient in a biobed. Moments later, the doctor received a comprehensive bio-scan report, including intricate details such as the patient’s DNA structure, amid many other medical information. In Star Trek, this is done in a matter of seconds, and based on these findings, an individualized gene therapy could be engineered in time to revert the disease. Although we have not seen anything similar to Star Trek’s medical tricorder yet, the last decades saw exponential growth in the understanding and application of genetics in diagnosing and managing human diseases. The completion of the Human Genome Project, a landmark international research project that mapped and sequenced the entire human genome, over 2 decades ago generated great optimism and expectation that genetic testing would drive a revolution in medical practice (1). Although some medical specialties, such as oncology, did see a rapid evolution in genetic diagnosis and management, the impact of genetics in critical care medicine has been more gradual. However, the development of faster and more cost-effective sequencing technologies is opening the genetic “revolution” for critical care patients. In neonatal ICUs, rapid whole genome sequencing (WGS) has been used to improve the capacity to diagnose numerous rare critical congenital diseases accurately. This approach has been shown to positively affect clinical management (2), and its potential as the next step in the evolution of neonatal screening is increasingly recognized. In PICUs, a similar approach to using WGS for rapid diagnosis of rare diseases has been used in smaller populations with promising results. In this issue of Pediatric Critical Care Medicine, Rodriguez et al (3) report the largest cohort so far to report WGS exclusively in PICUs. The authors describe the diagnostic yield and clinical utility of WGS in 133 children admitted to PICUs in four children’s hospitals, over 7 years. The study findings confirm previous data showing that genetic diagnoses are frequent in critically ill children and that results can be delivered in a clinically acceptable timeframe. The authors report an excellent 59% genetic diagnosis rate, in keeping with smaller PICU genomics cohorts from Taiwan (4) (52%) and Australia (5) (47%), and higher than (the still good yields from) U.K. (6) (25%) and U.S. studies (7) (24%). The average preliminary result was given in 3 (interquartile range, IQR 2–4) days and final reports in 12 (IQR 8–15) days, which is short enough to be meaningful even in a critical care setting. Similar rapid turnaround has also been reported in other studies (5), and recent development of newer and streamlined processes such as phenotyping automation (8) and ultrarapid nanotube genome sequencing have enabled further reduction in the timeframe for a genetic diagnosis to less than one day (9). Importantly, a genetic diagnosis led to a change in clinical management (acute or long-term) in more than half of the patients. The study by Rodriguez et al (3) has significant strengths, such as a multi-site design and a relatively large sample size, allowing for more meaningful data analysis than previous studies. Its main limitations are the mostly retrospective (73% of the patients recruited) design, frequent singleton analysis, and the unspecific selection criteria for recruitment into the study. Establishing selection criteria for studies on early diagnosis of genetic diseases is difficult. Patients with easily recognizable genetic syndromes may not need WGS testing, and diagnosis can often be confirmed in a timely manner using other strategies. For rare monogenic diseases—the usual focus of WGS—phenotypes are often variable, incomplete, or not fully expressed early in life. Therefore, selection based on patient phenotype is unlikely to enrich diagnostic yield and may miss a number of essential patients. In their study, Rodriguez et al (3) used a regression model to identify variables associated with a WGS diagnosis and found that dysmorphic features and congenital heart disease had the strongest associations. Dysmorphic features were also described as a phenotypic group with a higher diagnostic rate by French et al (6), but lactic acidosis and cerebral palsy and/or epilepsy were the groups with the highest diagnostic rate. Although these high diagnostic rate groups can assist pediatric intensivists in identifying candidates for WGS testing, they should not be solely relied upon as definitive selection criteria. Accordingly, within the diverse population of critically ill children admitted to PICUs there is currently no established set of criteria that can consistently improve the diagnostic yield of WGS beyond a physician’s clinical suspicion. French et al (6) previously attempted to use a set of broad and inclusive recruitment criteria, but among PICU patients, “physician request” was the most common inclusion criterion. Addressing the question of which critically ill children should have WGS will require substantially larger datasets for robust analysis and interpretation. Because a clinician suspicion of a monogenic disease is the current standard for requesting WGS, it is interesting that Rodriguez et al (3) report a similar rate of diagnosis yield for an indication from geneticists and pediatric intensivists (69% vs. 57%, with a mean difference of 12% [95% CI, –6.9% to 28.2%]). Because processing and reporting genomic diagnoses can now be done in a very short timeframe, a physician’s suspicion of monogenic disease is likely the next rate-limiting step for genomic diagnosis in critically ill children. This reinforces the need for pediatric intensivists to understand basic genomics, its applications, and limitations, and consider the possible future applications of genomics in critical care. Current genomic practices in pediatric critical care use both whole exome sequencing (WES) and WGS. WES targets the protein-coding genes (exome) regions, while WGS covers the entire genome, including both coding and noncoding regions. Although WES has the advantage of being cheaper and still providing good diagnostic yields (5), its ability to detect variants in regions of the genome that are difficult to sequence is limited, with inferior performance in identifying simple and complex structural variants with break points in the noncoding region of the genome (10), all of which WGS can do. Therefore, having the option, of WGS is preferable due to the higher quality of data and bigger diagnostic power. Once sequencing is complete, analysis is complex and resource-intensive. A gene-agnostic approach to genetic testing is important in PICUs because clinical phenotypes of rare disorders can vary significantly in this population, and phenotype-based hypotheses can lead to loss of diagnoses (6). Using trio analysis (patient and both parents) enhances the identification of missense variants because benign familial missense variants and novel pathogenic missense mutations can be more easily distinguished (6,11). After analysis, pathogenic/likely pathogenic findings should be correlated with clinical phenotype, validated, and only then disclosed. The reporting of changes in a gene unrelated to the primary purpose of the testing (secondary findings [SF]) is controversial, with significant regional variation in practice and recommendations (7,11). Similarly, variants that have features suggestive of possible functional consequence but for which there is insufficient evidence for either a pathogenic or benign role (variants of unknown significance [VUS]) are most frequently not reported unless they are linked to the patient phenotype or can inform subsequent studies to provide more insight into pathogenicity. Findings of VUS and/or SF are often not disclosed in the PICU genomic research (6,7), and pediatric intensivists need to understand these concepts and their implications for patients and families before disclosing genomic results. We are entering an era where genomics will become widely available, and diagnosis of rare diseases will become accurate and timely. In critical care, this has now bridged from research to clinical practice. However, the future of genomic research in critical care has much more to offer. Pharmacogenomic studies will use genetic profiles to optimize prescriptions in intensive care, aiming to provide unique information about individual responses to each drug. This will lead to individualized drug selection based on patient metabolism, also guiding dosages, and adjustments based on genetic profile. This should also predict and prevent adverse drug reactions. Functional genomics research will link genome findings with particular patterns of response to critical illness, allowing us to break down classic PICU syndromic diagnoses (e.g., sepsis, acute respiratory distress syndrome) into mechanistically distinct phenotypes that can be used for precision medicine treatments. A similar approach may be used in many critical illnesses (e.g., trauma), where knowledge of a patient’s pattern of response to stress may guide the use of modulatory therapies. Gene therapy could be used to correct pathologic and high-risk genetic variations early, allowing full treatment or at least prevention of disease progression. We are only starting with genomics in the PICU. There is a very long and complex path until we can fully understand the information that is being made available to us now. It will be a hard but rewarding work to apply this knowledge into clinical practice. Perhaps by that time, WGS will be done in a small portable, and noninvasive device that can be easily applied during starship missions. This may seem impossible for now, but as Captain Picard once said, “Things are only impossible until they’re not.” Live long and prosper.
Answers are grounded in this paper and cite their sources.
No takes yet. Share an insight, caveat, or question.
Branco et al. (2024) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: