A Latin American single-center exploratory latent class analysis

OBJECTIVES

To identify distinct subphenotypes of severe early-onset pre-eclampsia in Latin America and analyze biomarker and hemodynamic trends between subphenotypes after hospital admission.

METHODS

A single-center prospective cohort study was conducted in Colombia. The latent class analysis identified subphenotypes using clinical variables, biomarkers, laboratory tests, and maternal hemodynamics. Class-defining variables were restricted to measurements at and 24 h after admission. Primary and secondary outcomes were severe maternal and perinatal complications.

RESULTS

Among 49 patients, two subphenotypes were identified: Subphenotype 1 (34.7%) had a higher likelihood of an sFlt-1/PlGF ratio ≤ 38, maternal age > 35, and low probability of TPR > 1400, CO <8, and IUGR; Subphenotype 2 (65.3%) had a low likelihood of an sFlt-1/PlGF ratio < 38, maternal age > 35, and high probability of TPR > 1400, CO <8, and IUGR. At 24 h postadmission, 64.7% of subphenotype 1 patients changed to subphenotype 2, while 25% of subphenotype 2 patients were reclassified as subphenotype 1. Subphenotype 1 displayed significant changes in CO and TPR, while subphenotype 2 did not. Maternal complications were more prevalent in subphenotype 2, with an odds ratio of 5.3 (95% CI: 1.3-22.0; P = 0.02), but no significant differences in severe neonatal complications were observed.

CONCLUSIONS

We identified two distinct subphenotypes in a Latin American cohort of patients with severe early-onset pre-eclampsia. Subphenotype 2, characterized by higher TPR, sFlt-1, and serum creatinine and lower CO and PlGF at admission, was associated with worse maternal outcomes and appeared less modifiable after in-hospital treatment.

Using Trajectories of Bedside Vital Signs to Identify COVID-19 Subphenotypes

BACKGROUND

Trajectories of bedside vital signs have been used to identify sepsis subphenotypes with distinct outcomes and treatment responses. The objective of this study was to validate the vitals trajectory model in a multicenter cohort of patients hospitalized with COVID-19 and to evaluate the clinical characteristics and outcomes of the resulting subphenotypes.

RESEARCH QUESTION

Can the trajectory of routine bedside vital signs identify COVID-19 subphenotypes with distinct clinical characteristics and outcomes?

STUDY DESIGN AND METHODS

The study included adult patients admitted with COVID-19 to four academic hospitals in the Emory Healthcare system between March 1, 2020, and May 31, 2022. Using a validated group-based trajectory model, we classified patients into previously defined vital sign trajectories using oral temperature, heart rate, respiratory rate, and systolic and diastolic BP measured in the first 8 h of hospitalization. Clinical characteristics, biomarkers, and outcomes were compared between subphenotypes. Heterogeneity of treatment effect to tocilizumab was evaluated.

RESULTS

The 7,065 patients with hospitalized COVID-19 were classified into four subphenotypes: group A (n = 1,429, 20%)-high temperature, heart rate, respiratory rate, and hypotensive; group B (1,454, 21%)-high temperature, heart rate, respiratory rate, and hypertensive; group C (2,996, 42%)-low temperature, heart rate, respiratory rate, and normotensive; and group D (1,186, 17%)-low temperature, heart rate, respiratory rate, and hypotensive. Groups A and D had higher ORs of mechanical ventilation, vasopressors, and 30-day inpatient mortality (P < .001). On comparing patients receiving tocilizumab (n = 55) with those who met criteria for tocilizumab but were admitted before its use (n = 461), there was significant heterogeneity of treatment effect across subphenotypes in the association of tocilizumab with 30-day mortality (P = .001).

INTERPRETATION

By using bedside vital signs available in even low-resource settings, we found novel subphenotypes associated with distinct manifestations of COVID-19, which could lead to preemptive and targeted treatments.

Joint Modeling of Social Determinants and Clinical Factors to Define Subphenotypes in Out-of-Hospital Cardiac Arrest Survival: Cluster Analysis

Background

Machine learning clustering offers an unbiased approach to better understand the interactions of complex social and clinical variables via integrative subphenotypes, an approach not studied in out-of-hospital cardiac arrest (OHCA).

Objective

We conducted a cluster analysis for a cohort of OHCA survivors to examine the association of clinical and social factors for mortality at 1 year.

Methods

We used a retrospective observational OHCA cohort identified from Medicare claims data, including area-level social determinants of health (SDOH) features and hospital-level data sets. We applied k-means clustering algorithms to identify subphenotypes of beneficiaries who had survived an OHCA and examined associations of outcomes by subphenotype.

Results

We identified 27,028 unique beneficiaries who survived to discharge after OHCA. We derived 4 distinct subphenotypes. Subphenotype 1 included a distribution of more urban, female, and Black beneficiaries with the least robust area-level SDOH measures and the highest 1-year mortality (2375/4417, 53.8%). Subphenotype 2 was characterized by a greater distribution of male, White beneficiaries and had the strongest zip code-level SDOH measures, with 1-year mortality at 49.9% (4577/9165). Subphenotype 3 had the highest rates of cardiac catheterization at 34.7% (1342/3866) and the greatest distribution with a driving distance to the index OHCA hospital from their primary residence >16.1 km at 85.4% (8179/9580); more were also discharged to a skilled nursing facility after index hospitalization. Subphenotype 4 had moderate median household income at US $51,659.50 (IQR US $41,295 to $67,081) and moderate to high median unemployment at 5.5% (IQR 4.2%-7.1%), with the lowest 1-year mortality (1207/3866, 31.2%). Joint modeling of these features demonstrated an increased hazard of death for subphenotypes 1 to 3 but not for subphenotype 4 when compared to reference.

Conclusions

We identified 4 distinct subphenotypes with differences in outcomes by clinical and area-level SDOH features for OHCA. Further work is needed to determine if individual or other SDOH domains are specifically tied to long-term survival after OHCA.

COVID-19 subphenotypes at hospital admission are associated with mortality: a cross-sectional study

BACKGROUND

We have an incomplete understanding of COVID-19 characteristics at hospital presentation and whether underlying subphenotypes are associated with clinical outcomes and therapeutic responses.

METHODS

For this cross-sectional study, we extracted electronic health data from adults hospitalized between 1 March and 30 August 2020 with a PCR-confirmed diagnosis of COVID-19 at five New York City Hospitals. We obtained clinical and laboratory data from the first 24 h of the patient's hospitalization. Treatment with tocilizumab and convalescent plasma was assessed over hospitalization. The primary outcome was mortality; secondary outcomes included intubation, intensive care unit (ICU) admission and length of stay (LOS). First, we employed latent class analysis (LCA) to identify COVID-19 subphenotypes on admission without consideration of outcomes and assigned each patient to a subphenotype. We then performed robust Poisson regression to examine associations between COVID-19 subphenotype assignment and outcome. We explored whether the COVID-19 subphenotypes had a differential response to tocilizumab and convalescent plasma therapies.

RESULTS

A total of 4620 patients were included. LCA identified six subphenotypes, which were distinct by level of inflammation, clinical and laboratory derangements and ranged from a hypoinflammatory subphenotype with the fewest derangements to a hyperinflammatory with multiorgan dysfunction subphenotypes. Multivariable regression analyses found differences in risk for mortality, intubation, ICU admission and LOS, as compared to the hypoinflammatory subphenotype. For example, in multivariable analyses the moderate inflammation with fever subphenotype had 3.29 times the risk of mortality (95% CI 2.05, 5.28), while the hyperinflammatory with multiorgan failure subphenotype had 17.87 times the risk of mortality (95% CI 11.56, 27.63), as compared to the hypoinflammatory subphenotype. Exploratory analyses suggested that subphenotypes may differential respond to convalescent plasma or tocilizumab therapy.

CONCLUSION

COVID-19 subphenotype at hospital admission may predict risk for mortality, ICU admission and intubation and differential response to treatment.KEY MESSAGEThis cross-sectional study of COVID patients admitted to the Mount Sinai Health System, identified six distinct COVID subphenotypes on admission. Subphenotypes correlated with ICU admission, intubation, mortality and differential response to treatment.

Trajectories of Host-Response Subphenotypes in Patients With COVID-19 Across the Spectrum of Respiratory Support

BACKGROUND

Hospitalized patients with severe COVID-19 follow heterogeneous clinical trajectories, requiring different levels of respiratory support and experiencing diverse clinical outcomes. Differences in host immune responses to SARS-CoV-2 infection may account for the heterogeneous clinical course, but we have limited data on the dynamic evolution of systemic biomarkers and related subphenotypes. Improved understanding of the dynamic transitions of host subphenotypes in COVID-19 may allow for improved patient selection for targeted therapies.

RESEARCH QUESTION

We examined the trajectories of host-response profiles in severe COVID-19 and evaluated their prognostic impact on clinical outcomes.

STUDY DESIGN AND METHODS

In this prospective observational study, we enrolled 323 inpatients with COVID-19 receiving different levels of baseline respiratory support: (1) low-flow oxygen (37%), (2) noninvasive ventilation (NIV) or high-flow oxygen (HFO; 29%), (3) invasive mechanical ventilation (27%), and (4) extracorporeal membrane oxygenation (7%). We collected plasma samples on enrollment and at days 5 and 10 to measure host-response biomarkers. We classified patients by inflammatory subphenotypes using two validated predictive models. We examined clinical, biomarker, and subphenotype trajectories and outcomes during hospitalization.

RESULTS

IL-6, procalcitonin, and angiopoietin 2 persistently were elevated in patients receiving higher levels of respiratory support, whereas soluble receptor of advanced glycation end products (sRAGE) levels displayed the inverse pattern. Patients receiving NIV or HFO at baseline showed the most dynamic clinical trajectory, with 24% eventually requiring intubation and exhibiting worse 60-day mortality than patients receiving invasive mechanical ventilation at baseline (67% vs 35%; P < .0001). sRAGE levels predicted NIV failure and worse 60-day mortality for patients receiving NIV or HFO, whereas IL-6 levels were predictive in all patients regardless of level of support (P < .01). Patients classified to a hyperinflammatory subphenotype at baseline (< 10%) showed worse 60-day survival (P < .0001) and 50% of them remained classified as hyperinflammatory at 5 days after enrollment.

INTERPRETATION

Longitudinal study of the systemic host response in COVID-19 revealed substantial and predictive interindividual variability influenced by baseline levels of respiratory support.

Integrated Analysis of Blood and Urine Biomarkers to Identify Acute Kidney Injury Subphenotypes and Associations With Long-term Outcomes

RATIONALE & OBJECTIVE

Acute kidney injury (AKI) is a heterogeneous clinical syndrome with varying causes, pathophysiology, and outcomes. We incorporated plasma and urine biomarker measurements to identify AKI subgroups (subphenotypes) more tightly linked to underlying pathophysiology and long-term clinical outcomes.

STUDY DESIGN

Multicenter cohort study.

SETTING & PARTICIPANTS

769 hospitalized adults with AKI matched with 769 without AKI, enrolled from December 2009 to February 2015 in the ASSESS-AKI Study.

PREDICTORS

29 clinical, plasma, and urinary biomarker parameters used to identify AKI subphenotypes.

OUTCOME

Composite of major adverse kidney events (MAKE) with a median follow-up period of 4.7 years.

ANALYTICAL APPROACH

Latent class analysis (LCA) and k-means clustering were applied to 29 clinical, plasma, and urinary biomarker parameters. Associations between AKI subphenotypes and MAKE were analyzed using Kaplan-Meier curves and Cox proportional hazard models.

RESULTS

Among 769 AKI patients both LCA and k-means identified 2 distinct AKI subphenotypes (classes 1 and 2). The long-term risk for MAKE was higher with class 2 (adjusted HR, 1.41 [95% CI, 1.08-1.84]; P=0.01) compared with class 1, adjusting for demographics, hospital level factors, and KDIGO stage of AKI. The higher risk of MAKE among class 2 was explained by a higher risk of long-term chronic kidney disease progression and dialysis. The top variables that were different between classes 1 and 2 included plasma and urinary biomarkers of inflammation and epithelial cell injury; serum creatinine ranked 20th out of the 29 variables for differentiating classes.

LIMITATIONS

A replication cohort with simultaneously collected blood and urine sampling in hospitalized adults with AKI and long-term outcomes was unavailable.

CONCLUSIONS

We identify 2 molecularly distinct AKI subphenotypes with differing risk of long-term outcomes, independent of the current criteria to risk stratify AKI. Future identification of AKI subphenotypes may facilitate linking therapies to underlying pathophysiology to prevent long-term sequalae after AKI.

PLAIN-LANGUAGE SUMMARY

Acute kidney injury (AKI) occurs commonly in hospitalized patients and is associated with high morbidity and mortality. The AKI definition lumps many different types of AKI together, but subgroups of AKI may be more tightly linked to the underlying biology and clinical outcomes. We used 29 different clinical, blood, and urinary biomarkers and applied 2 different statistical algorithms to identify AKI subtypes and their association with long-term outcomes. Both clustering algorithms identified 2 AKI subtypes with different risk of chronic kidney disease, independent of the serum creatinine concentrations (the current gold standard to determine severity of AKI). Identification of AKI subtypes may facilitate linking therapies to underlying biology to prevent long-term consequences after AKI.

Cardiovascular Subphenotypes in ARDS: Diagnostic and Therapeutic Implications and Overlap with Other ARDS Subphenotypes

Acute respiratory distress syndrome (ARDS) is a highly heterogeneous clinical condition. Shock is a poor prognostic sign in ARDS, and heterogeneity in its pathophysiology may be a barrier to its effective treatment. Although right ventricular dysfunction is commonly implicated, there is no consensus definition for its diagnosis, and left ventricular function is neglected. There is a need to identify the homogenous subgroups within ARDS, that have a similar pathobiology, which can then be treated with targeted therapies. Haemodynamic clustering analyses in patients with ARDS have identified two subphenotypes of increasingly severe right ventricular injury, and a further subphenotype of hyperdynamic left ventricular function. In this review, we discuss how phenotyping the cardiovascular system in ARDS may align with haemodynamic pathophysiology, can aid in optimally defining right ventricular dysfunction and can identify tailored therapeutic targets for shock in ARDS. Additionally, clustering analyses of inflammatory, clinical and radiographic data describe other subphenotypes in ARDS. We detail the potential overlap between these and the cardiovascular phenotypes.

Comparison of time series clustering methods for identifying novel subphenotypes of patients with infection

OBJECTIVE

Severe infection can lead to organ dysfunction and sepsis. Identifying subphenotypes of infected patients is essential for personalized management. It is unknown how different time series clustering algorithms compare in identifying these subphenotypes.

MATERIALS AND METHODS

Patients with suspected infection admitted between 2014 and 2019 to 4 hospitals in Emory healthcare were included, split into separate training and validation cohorts. Dynamic time warping (DTW) was applied to vital signs from the first 8 h of hospitalization, and hierarchical clustering (DTW-HC) and partition around medoids (DTW-PAM) were used to cluster patients into subphenotypes. DTW-HC, DTW-PAM, and a previously published group-based trajectory model (GBTM) were evaluated for agreement in subphenotype clusters, trajectory patterns, and subphenotype associations with clinical outcomes and treatment responses.

RESULTS

There were 12 473 patients in training and 8256 patients in validation cohorts. DTW-HC, DTW-PAM, and GBTM models resulted in 4 consistent vitals trajectory patterns with significant agreement in clustering (71-80% agreement, P < .001): group A was hyperthermic, tachycardic, tachypneic, and hypotensive. Group B was hyperthermic, tachycardic, tachypneic, and hypertensive. Groups C and D had lower temperatures, heart rates, and respiratory rates, with group C normotensive and group D hypotensive. Group A had higher odds ratio of 30-day inpatient mortality (P < .01) and group D had significant mortality benefit from balanced crystalloids compared to saline (P < .01) in all 3 models.

DISCUSSION

DTW- and GBTM-based clustering algorithms applied to vital signs in infected patients identified consistent subphenotypes with distinct clinical outcomes and treatment responses.

CONCLUSION

Time series clustering with distinct computational approaches demonstrate similar performance and significant agreement in the resulting subphenotypes.

Identifying acute kidney injury subphenotypes using an outcome-driven deep-learning approach

OBJECTIVE

Acute kidney injury (AKI), a common condition on the intensive-care unit (ICU), is characterized by an abrupt decrease in kidney function within a few hours or days, leading to kidney failure or damage. Although AKI is associated with poor outcomes, current guidelines overlook the heterogeneity among patients with this condition. Identification of AKI subphenotypes could enable targeted interventions and a deeper understanding of the injury's pathophysiology. While previous approaches based on unsupervised representation learning have been used to identify AKI subphenotypes, these methods cannot assess time series or disease severity.

METHODS

In this study, we developed a data- and outcome-driven deep-learning (DL) approach to identify and analyze AKI subphenotypes with prognostic and therapeutic implications. Specifically, we developed a supervised long short-term memory (LSTM) autoencoder (AE) with the aim of extracting representation from time-series EHR data that were intricately correlated with mortality. Then, subphenotypes were identified via application of K-means.

RESULTS

In two publicly available datasets, three distinct clusters were identified, characterized by mortality rates of 11.3%, 17.3%, and 96.2% in one dataset and 4.6%, 12.1%, and 54.6% in the other. Further analysis demonstrated that AKI subphenotypes identified by our proposed approach were statistically significant on several clinical characteristics and outcomes.

CONCLUSION

In this study, our proposed approach could successfully cluster the AKI population in ICU settings into 3 distinct subphenotypes. Thus, such approach could potentially improve outcomes of AKI patients in the ICU, with better risk assessment and potentially better personalized treatment.

The Gut Microbiome Modulates Body Temperature Both in Sepsis and Health

Rationale: Among patients with sepsis, variation in temperature trajectories predicts clinical outcomes. In healthy individuals, normal body temperature is variable and has decreased consistently since the 1860s. The biologic underpinnings of this temperature variation in disease and health are unknown. Objectives: To establish and interrogate the role of the gut microbiome in calibrating body temperature. Methods: We performed a series of translational analyses and experiments to determine whether and how variation in gut microbiota explains variation in body temperature in sepsis and in health. We studied patient temperature trajectories using electronic medical record data. We characterized gut microbiota in hospitalized patients using 16S ribosomal RNA gene sequencing. We modeled sepsis using intraperitoneal LPS in mice and modulated the microbiome using antibiotics, germ-free, and gnotobiotic animals. Measurements and Main Results: Consistent with prior work, we identified four temperature trajectories in patients hospitalized with sepsis that predicted clinical outcomes. In a separate cohort of 116 hospitalized patients, we found that the composition of patients' gut microbiota at admission predicted their temperature trajectories. Compared with conventional mice, germ-free mice had reduced temperature loss during experimental sepsis. Among conventional mice, heterogeneity of temperature response in sepsis was strongly explained by variation in gut microbiota. Healthy germ-free and antibiotic-treated mice both had lower basal body temperatures compared with control animals. The Lachnospiraceae family was consistently associated with temperature trajectories in hospitalized patients, experimental sepsis, and antibiotic-treated mice. Conclusions: The gut microbiome is a key modulator of body temperature variation in both health and critical illness and is thus a major, understudied target for modulating physiologic heterogeneity in sepsis.

Evolution of Clinical Phenotypes of COVID-19 Patients During Intensive Care Treatment: An Unsupervised Machine Learning Analysis

BACKGROUND

Identification of clinical phenotypes in critically ill COVID-19 patients could improve understanding of the disease heterogeneity and enable prognostic and predictive enrichment. However, previous attempts did not take into account temporal dynamics with high granularity. By including the dimension of time, we aim to gain further insights into the heterogeneity of COVID-19.

METHODS

We used granular data from 3202 adult COVID patients in the Dutch Data Warehouse that were admitted to one of 25 Dutch ICUs between February 2020 and March 2021. Parameters including demographics, clinical observations, medications, laboratory values, vital signs, and data from life support devices were selected. Twenty-one datasets were created that each covered 24 h of ICU data for each day of ICU treatment. Clinical phenotypes in each dataset were identified by performing cluster analyses. Both evolution of the clinical phenotypes over time and patient allocation to these clusters over time were tracked.

RESULTS

The final patient cohort consisted of 2438 COVID-19 patients with a ICU mortality outcome. Forty-one parameters were chosen for cluster analysis. On admission, both a mild and a severe clinical phenotype were found. After day 4, the severe phenotype split into an intermediate and a severe phenotype for 11 consecutive days. Heterogeneity between phenotypes appears to be driven by inflammation and dead space ventilation. During the 21-day period, only 8.2% and 4.6% of patients in the initial mild and severe clusters remained assigned to the same phenotype respectively. The clinical phenotype half-life was between 5 and 6 days for the mild and severe phenotypes, and about 3 days for the medium severe phenotype.

CONCLUSIONS

Patients typically do not remain in the same cluster throughout intensive care treatment. This may have important implications for prognostic or predictive enrichment. Prominent dissimilarities between clinical phenotypes are predominantly driven by inflammation and dead space ventilation.

Biological Phenotyping in Sepsis and Acute Respiratory Distress Syndrome

Heterogeneity in sepsis and acute respiratory distress syndrome (ARDS) is increasingly being recognized as one of the principal barriers to finding efficacious targeted therapies. The advent of multiple high-throughput biological data ("omics"), coupled with the widespread access to increased computational power, has led to the emergence of phenotyping in critical care. Phenotyping aims to use a multitude of data to identify homogenous subgroups within an otherwise heterogenous population. Increasingly, phenotyping schemas are being applied to sepsis and ARDS to increase understanding of these clinical conditions and identify potential therapies. Here we present a selective review of the biological phenotyping schemas applied to sepsis and ARDS. Further, we outline some of the challenges involved in translating these conceptual findings to bedside clinical decision-making tools.

Identification of subphenotypes in critically ill thrombocytopenic patients with different responses to therapeutic interventions: a retrospective study

Introduction

The causes of thrombocytopenia (TP) in critically ill patients are numerous and heterogeneous. Currently, subphenotype identification is a popular approach to address this problem. Therefore, this study aimed to identify subphenotypes that respond differently to therapeutic interventions in patients with TP using routine clinical data and to improve individualized management of TP.

Methods

This retrospective study included patients with TP admitted to the intensive care unit (ICU) of Dongyang People's Hospital during 2010-2020. Subphenotypes were identified using latent profile analysis of 15 clinical variables. The Kaplan-Meier method was used to assess the risk of 30-day mortality for different subphenotypes. Multifactorial Cox regression analysis was used to analyze the relationship between therapeutic interventions and in-hospital mortality for different subphenotypes.

Results

This study included a total of 1,666 participants. Four subphenotypes were identified by latent profile analysis, with subphenotype 1 being the most abundant and having a low mortality rate. Subphenotype 2 was characterized by respiratory dysfunction, subphenotype 3 by renal insufficiency, and subphenotype 4 by shock-like features. Kaplan-Meier analysis revealed that the four subphenotypes had different in-30-day mortality rates. The multivariate Cox regression analysis indicated a significant interaction between platelet transfusion and subphenotype, with more platelet transfusion associated with a decreased risk of in-hospital mortality in subphenotype 3 [hazard ratio (HR): 0.66, 95% confidence interval (CI): 0.46-0.94]. In addition, there was a significant interaction between fluid intake and subphenotype, with a higher fluid intake being associated with a decreased risk of in-hospital mortality for subphenotype 3 (HR: 0.94, 95% CI: 0.89-0.99 per 1 l increase in fluid intake) and an increased risk of in-hospital mortality for high fluid intake in subphenotypes 1 (HR: 1.10, 95% CI: 1.03-1.18 per 1 l increase in fluid intake) and 2 (HR: 1.19, 95% CI: 1.08-1.32 per 1 l increase in fluid intake).

Conclusion

Four subphenotypes of TP in critically ill patients with different clinical characteristics and outcomes and differential responses to therapeutic interventions were identified using routine clinical data. These findings can help improve the identification of different subphenotypes in patients with TP for better individualized treatment of patients in the ICU.

Trajectories of host-response biomarkers and inflammatory subphenotypes in COVID-19 patients across the spectrum of respiratory support

Purpose

Enhanced understanding of the dynamic changes in the dysregulated inflammatory response in COVID-19 may help improve patient selection and timing for immunomodulatory therapies.

Methods

We enrolled 323 COVID-19 inpatients on different levels of baseline respiratory support: i) Low Flow Oxygen (37%), ii) Non-Invasive Ventilation or High Flow Oxygen (NIV_HFO, 29%), iii) Invasive Mechanical Ventilation (IMV, 27%), and iv) Extracorporeal Membrane Oxygenation (ECMO, 7%). We collected plasma samples upon enrollment and days 5 and 10 to measure host-response biomarkers. We classified subjects into inflammatory subphenotypes using two validated predictive models. We examined clinical, biomarker and subphenotype trajectories and outcomes during hospitalization.

Results

IL-6, procalcitonin, and Angiopoietin-2 were persistently elevated in patients at higher levels of respiratory support, whereas sRAGE displayed the inverse pattern. Patients on NIV_HFO at baseline had the most dynamic clinical trajectory, with 26% eventually requiring intubation and exhibiting worse 60-day mortality than IMV patients at baseline (67% vs. 35%, p<0.0001). sRAGE levels predicted NIV failure and worse 60-day mortality for NIV_HFO patients, whereas IL-6 levels were predictive in IMV or ECMO patients. Hyper-inflammatory subjects at baseline (<10% by both models) had worse 60-day survival (p<0.0001) and 50% of them remained classified as hyper-inflammatory on follow-up sampling at 5 days post-enrollment. Receipt of combined immunomodulatory therapies (steroids and anti-IL6 agents) was associated with markedly increased IL-6 and lower Angiopoietin-2 levels (p<0.05).

Conclusions

Longitudinal study of systemic host responses in COVID-19 revealed substantial and predictive inter-individual variability, influenced by baseline levels of respiratory support and concurrent immunomodulatory therapies.

Clinical, Virological and Immunological Subphenotypes in a Cohort of Early Treated HIV-Infected Children

Background

Identifying subphenotypes within heterogeneous diseases may have an impact in terms of therapeutic options. In this study, we aim to assess different subphenotypes in children living with human immunodeficiency virus (HIV-1), according to the clinical, virological, and immunological characteristics.

Methods

We collected clinical and sociodemographic data, baseline viral load (VL), CD4 and CD8 count and percentage, age at initiation of ART, HIV DNA reservoir size in peripheral blood mononuclear cells (PBMCs), cell-associated RNA (CA-RNA), ultrasensitive VL, CD4 subsets (T effector CD25+, activated memory cells, Treg cells), humoral-specific HIV response (T-bet B cells), innate response (CD56dim natural killer (NK) cells, NKp46+, perforin), exhaustion markers (PD-1, PD-L1, DNAM), CD8 senescence, and biomarkers for T-lymphocyte thymic output (TREC) and endothelial activation (VCAM). The most informative variables were selected using an unsupervised lasso-type penalty selection for sparse clustering. Hierarchical clustering was performed using Pearson correlation as the distance metric and WARD.D2 as the clustering method. Internal validation was applied to select the best number of clusters. To compare the characteristics among clusters, boxplot and Kruskal Wallis test were assessed.

Results

Three subphenotypes were discovered (cluster1: n=18, 45%; cluster2: n=11, 27.5%; cluster3: n=11, 27.5%). Patients in cluster1 were treated earlier, had higher baseline %CD4, low HIV reservoir size, low western blot score, higher TREC values, and lower VCAM values than the patients in the other clusters. In contrast, cluster3 was the less favorable. Patients were treated later and presented poorer outcomes with lower %CD4, and higher reservoir size, along with a higher percentage of CD8 immunosenescent cells, lower TREC, higher VCAM cytokine, and a higher %CD4 PD-1. Cluster2 was intermediate. Patients were like those of cluster1, but had lower levels of t-bet expression and higher HIV DNA reservoir size.

Conclusions

Three HIV pediatric subphenotypes with different virological and immunological features were identified. The most favorable cluster was characterized by a higher rate of immune reconstitution and a slower disease progression, and the less favorable with more senescence and high reservoir size. In the near future therapeutic interventions for a path of a cure might be guided or supported by the different subphenotypes.

Single Cell Scale Neuronal and Glial Gene Expression and Putative Cell Phenotypes and Networks in the Nucleus Tractus Solitarius in an Alcohol Withdrawal Time Series

Alcohol withdrawal syndrome (AWS) is characterized by neuronal hyperexcitability, autonomic dysregulation, and severe negative emotion. The nucleus tractus solitarius (NTS) likely plays a prominent role in the neurological processes underlying these symptoms as it is the main viscerosensory nucleus in the brain. The NTS receives visceral interoceptive inputs, influences autonomic outputs, and has strong connections to the limbic system and hypothalamic-pituitary-adrenal axis to maintain homeostasis. Our prior analysis of single neuronal gene expression data from the NTS shows that neurons exist in heterogeneous transcriptional states that form distinct functional subphenotypes. Our working model conjectures that the allostasis secondary to alcohol dependence causes peripheral and central biological network decompensation in acute abstinence resulting in neurovisceral feedback to the NTS that substantially contributes to the observed AWS. We collected single noradrenergic and glucagon-like peptide-1 (GLP-1) neurons and microglia from rat NTS and measured a subset of their transcriptome as pooled samples in an alcohol withdrawal time series. Inflammatory subphenotypes predominate at certain time points, and GLP-1 subphenotypes demonstrated hyperexcitability post-withdrawal. We hypothesize such inflammatory and anxiogenic signaling contributes to alcohol dependence via negative reinforcement. Targets to mitigate such dysregulation and treat dependence can be identified from this dataset.

Coagulation Dysfunction in Acute Respiratory Distress Syndrome and Its Potential Impact in Inflammatory Subphenotypes

The Acute Respiratory Distress Syndrome (ARDS) has caused innumerable deaths worldwide since its initial description over five decades ago. Population-based estimates of ARDS vary from 1 to 86 cases per 100,000, with the highest rates reported in Australia and the United States. This syndrome is characterised by a breakdown of the pulmonary alveolo-epithelial barrier with subsequent severe hypoxaemia and disturbances in pulmonary mechanics. The underlying pathophysiology of this syndrome is a severe inflammatory reaction and associated local and systemic coagulation dysfunction that leads to pulmonary and systemic damage, ultimately causing death in up to 40% of patients. Since inflammation and coagulation are inextricably linked throughout evolution, it is biological folly to assess the two systems in isolation when investigating the underlying molecular mechanisms of coagulation dysfunction in ARDS. Although the body possesses potent endogenous systems to regulate coagulation, these become dysregulated and no longer optimally functional during the acute phase of ARDS, further perpetuating coagulation, inflammation and cell damage. The inflammatory ARDS subphenotypes address inflammatory differences but neglect the equally important coagulation pathway. A holistic understanding of this syndrome and its subphenotypes will improve our understanding of underlying mechanisms that then drive translation into diagnostic testing, treatments, and improve patient outcomes.

Biomarker-Based Classification of Patients With Acute Respiratory Failure Into Inflammatory Subphenotypes: A Single-Center Exploratory Study

OBJECTIVES

Hyper- and hypoinflammatory subphenotypes discovered in patients with acute respiratory distress syndrome predict clinical outcomes and therapeutic responses. These subphenotypes may be important in broader critically ill patient populations with acute respiratory failure regardless of clinical diagnosis. We investigated subphenotyping with latent class analysis in an inclusive population of acute respiratory failure, derived a parsimonious model for subphenotypic predictions based on a small set of variables, and examined associations with clinical outcomes.

DESIGN

Prospective, observational cohort study.

SETTING

Single-center, academic medical ICU.

PATIENTS

Mechanically ventilated patients with acute respiratory failure.

MEASUREMENTS AND MAIN RESULTS

We included 498 patients with acute respiratory failure (acute respiratory distress syndrome: 143, at-risk for acute respiratory distress syndrome: 198, congestive heart failure: 37, acute on chronic respiratory failure: 23, airway protection: 61, and multifactorial: 35) in our derivation cohort and measured 10 baseline plasma biomarkers. Latent class analysis considering clinical variables and biomarkers determined that a two-class model offered optimal fit (23% hyperinflammatory subphenotype). Distribution of hyperinflammatory subphenotype varied among acute respiratory failure etiologies (acute respiratory distress syndrome: 31%, at-risk for acute respiratory distress syndrome: 27%, congestive heart failure: 22%, acute on chronic respiratory failure 0%, airway protection: 5%, and multifactorial: 14%). Hyperinflammatory patients had higher Sequential Organ Failure Assessment scores, fewer ventilator-free days, and higher 30- and 90-day mortality (all p < 0.001). We derived a parsimonious model consisting of angiopoietin-2, soluble tumor necrosis factor receptor-1, procalcitonin, and bicarbonate and classified subphenotypes in a validation cohort (n = 139). Hyperinflammatory patients (19%) demonstrated higher levels of inflammatory biomarkers not included in the model (p < 0.01) and worse outcomes.

CONCLUSIONS

Host-response subphenotypes are observable in a heterogeneous population with acute respiratory failure and predict clinical outcomes. Simple, biomarker-based models can offer prognostic enrichment in patients with acute respiratory failure. The differential distribution of subphenotypes by specific etiologies of acute respiratory failure indicates that subphenotyping may be more relevant in patients with hypoxemic causes of acute respiratory failure and not in patients intubated for airway protection or acute on chronic decompensation.

Identification of Distinct Clinical Subphenotypes in Critically Ill Patients With COVID-19

Background

Subphenotypes have been identified in patients with sepsis and ARDS and are associated with different outcomes and responses to therapies.

Research Question

Can unique subphenotypes be identified among critically ill patients with COVID-19?

Study Design and Methods

Using data from a multicenter cohort study that enrolled critically ill patients with COVID-19 from 67 hospitals across the United States, we randomly divided centers into discovery and replication cohorts. We used latent class analysis independently in each cohort to identify subphenotypes based on clinical and laboratory variables. We then analyzed the associations of subphenotypes with 28-day mortality.

Results

Latent class analysis identified four subphenotypes (SP) with consistent characteristics across the discovery (45 centers; n = 2,188) and replication (22 centers; n = 1,112) cohorts. SP1 was characterized by shock, acidemia, and multiorgan dysfunction, including acute kidney injury treated with renal replacement therapy. SP2 was characterized by high C-reactive protein, early need for mechanical ventilation, and the highest rate of ARDS. SP3 showed the highest burden of chronic diseases, whereas SP4 demonstrated limited chronic disease burden and mild physiologic abnormalities. Twenty-eight-day mortality in the discovery cohort ranged from 20.6% (SP4) to 52.9% (SP1). Mortality across subphenotypes remained different after adjustment for demographics, comorbidities, organ dysfunction and illness severity, regional and hospital factors. Compared with SP4, the relative risks were as follows: SP1, 1.67 (95% CI, 1.36-2.03); SP2, 1.39 (95% CI, 1.17-1.65); and SP3, 1.39 (95% CI, 1.15-1.67). Findings were similar in the replication cohort.

Interpretation

We identified four subphenotypes of COVID-19 critical illness with distinct patterns of clinical and laboratory characteristics, comorbidity burden, and mortality.

Phenotypes and Subphenotypes of Patients With COVID-19: A Latent Class Modeling Analysis

Background

Since COVID-19 was identified, its clinical and biological heterogeneity has been recognized. Identifying COVID-19 phenotypes might help guide basic, clinical, and translational research efforts.

Research Question

Does the clinical spectrum of patients with COVID-19 contain distinct phenotypes and subphenotypes?

Study Design and Methods

We included adult patients (≥ 18 years) positive for laboratory-confirmed SARS-CoV-2 infection from a prospective COVID-19 registry database in the Cleveland Clinic Health System in Ohio and Florida. The patients were split into training and testing sets. Using latent class analysis (LCA), we first identified phenotypic clusters of patients with COVID-19 based on demographics, comorbidities, and presenting symptoms. We then identified subphenotypes of hospitalized patients with additional blood biomarker data measured on hospital admission. The associations of phenotypes/subphenotypes and clinical outcomes were investigated. Multivariable prediction models were established to predict assignment to the LCA-defined phenotypes and subphenotypes and then evaluated on an independent testing set.

Results

We analyzed data for 20,572 patients. Seven phenotypes were identified on the basis of different profiles of presenting COVID-19 symptoms and existing comorbidities, including the following groups: young, no symptoms; young, symptoms; middle-aged, no symptoms; middle-aged, symptoms; middle-aged, comorbidities; old, no symptoms; and old, symptoms. The rates of inpatient hospitalization for the phenotypes were significantly different (P < .001). Five subphenotypes were identified for the subgroup of hospitalized patients, including the following subgroups: young, elevated WBC and platelet counts; middle-aged, lymphopenic with elevated C-reactive protein; middle-aged, hyperinflammatory; old, leukopenic with comorbidities; and old, hyperinflammatory with kidney dysfunction. The hospital mortality and the times from hospitalization to ICU transfer or death were significantly different (P < .001). The models for predicting the LCA-defined phenotypes and subphenotypes showed high discrimination (concordance index, 0.92 and 0.91).

Interpretation

Hypothesis-free LCA-defined phenotypes and subphenotypes of patients with COVID-19 can be identified. These may help clinical investigators conduct stratified analyses in clinical trials and assist basic science researchers in characterizing the pathobiology of the spectrum of COVID-19 presentations.

Seasonal Influence on the Numbers of Gender-Related Orofacial Cleft Conceptions in the Netherlands

BACKGROUND

In the multifactorial etiology of orofacial clefts (OFCs), environmental factors play an important role. To trace the influence of these factors, the timing of the cell biological mechanisms that occur during embryological development of the primary and secondary palates must be taken into account. That is, the fusion process of the facial and palatal processes, respectively, followed by their differentiation into bone and musculature, which take place during the first trimester of pregnancy. During this period, harmful seasonal influences such as viral infections and vitamin deficiencies could induce OFC in the embryo.

AIMS

The aim of this study is to find out whether a seasonal conception period with an increased risk of OFC development exists, particularly gender related.

METHODS

This was a retrospective cross-sectional study on children with OFC born in the Netherlands from 2006 to 2016. Total conception rates of live births in the Netherlands were used as a control group. χ² tests were performed to analyze monthly and seasonal differences. Males and females, positive and negative family history and subphenotype groups based on fusion and/or differentiation (F- and/or D-) defects, and their timing in embryogenesis were analyzed separately.

RESULTS

In total, 1653 children with OFC, 1041 males and 612 females, were analyzed. Only males with FD-defects showed a significant seasonal variation with an increase in conceptions during spring, most often in May.

CONCLUSIONS

Males with FD-defects showed a significant seasonal variation with an increase in conceptions during spring. No other seasonal trends could be demonstrated.

Data-driven discovery of probable Alzheimer's disease and related dementia subphenotypes using electronic health records

Introduction

We sought to assess longitudinal electronic health records (EHRs) using machine learning (ML) methods to computationally derive probable Alzheimer's Disease (AD) and related dementia subphenotypes.

Methods

A retrospective analysis of EHR data from a cohort of 7587 patients seen at a large, multi‐specialty urban academic medical center in New York was conducted. Subphenotypes were derived using hierarchical clustering from 792 probable AD patients (cases) who had received at least one diagnosis of AD using their clinical data. The other 6795 patients, labeled as controls, were matched on age and gender with the cases and randomly selected in the ratio of 9:1. Prediction models with multiple ML algorithms were trained on this cohort using 5‐fold cross‐validation. XGBoost was used to rank the variable importance.

Results

Four subphenotypes were computationally derived. Subphenotype A (n = 273; 28.2%) had more patients with cardiovascular diseases; subphenotype B (n = 221; 27.9%) had more patients with mental health illnesses, such as depression and anxiety; patients in subphenotype C (n = 183; 23.1%) were overall older (mean (SD) age, 79.5 (5.4) years) and had the most comorbidities including diabetes, cardiovascular diseases, and mental health disorders; and subphenotype D (n = 115; 14.5%) included patients who took anti‐dementia drugs and had sensory problems, such as deafness and hearing impairment.

The 0‐year prediction model for AD risk achieved an area under the receiver operating curve (AUC) of 0.764 (SD: 0.02); the 6‐month model, 0.751 (SD: 0.02); the 1‐year model, 0.752 (SD: 0.02); the 2‐year model, 0.749 (SD: 0.03); and the 3‐year model, 0.735 (SD: 0.03), respectively. Based on variable importance, the top‐ranked comorbidities included depression, stroke/transient ischemic attack, hypertension, anxiety, mobility impairments, and atrial fibrillation. The top‐ranked medications included anti‐dementia drugs, antipsychotics, antiepileptics, and antidepressants.

Conclusions

Four subphenotypes were computationally derived that correlated with cardiovascular diseases and mental health illnesses. ML algorithms based on patient demographics, diagnosis, and treatment demonstrated promising results in predicting the risk of developing AD at different time points across an individual's lifespan.

Embryologically Based Classification Specifies Gender Differences in the Prevalence of Orofacial Cleft Subphenotypes

Background:

A recently published validated classification system divides all orofacial cleft (OFC) subphenotypes into groups based on underlying developmental mechanisms, that is, fusion and differentiation, and their timing, that is, early and late periods, in embryogenesis of the primary and secondary palates.

Aims:

The aim of our study was to define gender differences in prevalence for all subphenotypes in newborns with OFC in the Netherlands.

Methods:

This was a retrospective cross-sectional study on children with OFC born from 2006 to 2016. Clefts were classified in early (E-), late (L-), and early/late (EL-) embryonic periods, in primary (P-), secondary (S-), and primary/secondary (PS-) palates, and further divided into fusion (F-), differentiation (D-), and fusion/differentiation (FD-) defects, respectively.

Results:

A total of 2089 OFC children were analyzed (1311 males and 778 females). Orofacial cleft subphenotypes in females occurred significantly more frequent in the L-period compared to males (66% vs 55%, P = .000), whereas clefts in males occurred significantly more in the EL-periods (40% vs 27%, P = .000). Females had significantly more S-palatal clefts (42% vs 23%, P = .000), while males had significantly more PS-palatal clefts (44% vs 30%, P = .000). Furthermore, the clefts in females were significantly more frequent the result of an F-defect (60% vs 52%, P = .000).

Conclusions:

Orofacial cleft in females mainly occur in the L-period are mostly S-palatal clefts, and are usually the result of an F-defect. Orofacial cleft in males more commonly occur in the EL-periods, are therefore more often combined PS-palatal clefts, and are more frequent D- and FD-defects.

Identification of Acute Kidney Injury Subphenotypes with Differing Molecular Signatures and Responses to Vasopressin Therapy

RATIONALE

Currently, no safe and effective pharmacologic interventions exist for acute kidney injury (AKI). One reason may be that heterogeneity exists within the AKI population, thereby hampering the identification of specific pathophysiologic pathways and therapeutic targets.

OBJECTIVE

The aim of this study was to identify and test whether AKI subphenotypes have prognostic and therapeutic implications.

METHODS

First, latent class analysis methodology was applied independently in two critically ill populations (discovery [n = 794] and replication [n = 425]) with AKI. Second, a parsimonious classification model was developed to identify AKI subphenotypes. Third, the classification model was applied to patients with AKI in VASST (Vasopressin and Septic Shock Trial; n = 271), and differences in treatment response were determined. In all three populations, AKI was defined using serum creatinine and urine output.

MEASUREMENTS AND MAIN RESULTS

A two-subphenotype latent class analysis model had the best fit in both the discovery (P = 0.004) and replication (P = 0.004) AKI groups. The risk of 7-day renal nonrecovery and 28-day mortality was greater with AKI subphenotype 2 (AKI-SP2) relative to AKI subphenotype 1 (AKI-SP1). The AKI subphenotypes discriminated risk for poor clinical outcomes better than the Kidney Disease: Improving Global Outcomes stages of AKI. A three-variable model that included markers of endothelial dysfunction and inflammation accurately determined subphenotype membership (C-statistic 0.92). In VASST, vasopressin compared with norepinephrine was associated with improved 90-day mortality in AKI-SP1 (27% vs. 46%, respectively; P = 0.02), but no significant difference was observed in AKI-SP2 (45% vs. 49%, respectively; P = 0.99) and the P value for interaction was 0.05.

CONCLUSIONS

This analysis identified two molecularly distinct AKI subphenotypes with different clinical outcomes and responses to vasopressin therapy. Identification of AKI subphenotypes could improve risk prognostication and may be useful for predictive enrichment in clinical trials.

Rare variant association testing for multicategory phenotype

Genetic association studies have provided new insights into the genetic variability of human complex traits with a focus mainly on continuous or binary traits. Methods have been proposed to take into account disease heterogeneity between subgroups of patients when studying common variants but none was specifically designed for rare variants. Because rare variants are expected to have stronger effects and to be more heterogeneously distributed among cases than common ones, subgroup analyses might be particularly attractive in this context. To address this issue, we propose an extension of burden tests by using a multinomial regression model, which enables association tests between rare variants and multicategory phenotypes. We evaluated the type I error and the power of two burden tests, CAST and WSS, by simulating data under different scenarios. In the case of genetic heterogeneity between case subgroups, we showed an advantage of multinomial regression over logistic regression, which considers all the cases against the controls. We replicated these results on real data from Moyamoya disease where the burden tests performed better when cases were stratified according to age-of-onset. We implemented the functions for association tests in the R package "Ravages" available on Github.

Insistence on sameness and broader autism phenotype in simplex families with autism spectrum disorder

Insistence on sameness (IS) in individuals with autism spectrum disorder (ASD) and their families may have utility in identifying meaningful subgroups for studying the pathophysiological and genetic pathways affected in ASD. The primary objectives of the current study were to (1) characterize features of IS in parents of children with ASD and (2) examine their relationships with child IS symptoms. Participants were 2760 families who participated in the Simons Simplex Collection. Levels of parent IS were measured using the Broader Autism Phenotype Questionnaire (BAPQ). A factor analysis generated a BAPQ-IS scale, consisting of a subset of 11 items from the original BAPQ-Rigid scale. Correlations were run to examine the relationship between parent BAP and child IS variables. Correlations were found between parent IS and measures of child IS. Although relationships between parent and child IS features were statistically significant in this large sample, effect sizes were small. Results may be reflective of sample design that only included simplex families, where ASD severity may be predominantly driven by spontaneous mutations and less by common inherited risk from parents. In addition, child and parent measures used may have differentially captured features and severity of IS. Further research is needed on how IS can be accurately measured throughout development and across individuals with ASD and their unaffected family members to facilitate future studies on IS as a possible endophenotype for ASD. Autism Res 2018, 11: 1253-1263. © 2018 International Society for Autism Research, Wiley Periodicals, Inc. LAY SUMMARY: Previous research has suggested that insistence on sameness (IS) may be a heritable trait in autism spectrum disorder (ASD). The study examined whether children with high levels of IS had parents with IS tendencies. A small relationship was found between parent and child measures of IS. Future research is needed on measurement of insistence on sameness across individuals with and without ASD to further examine this relationship and improve understanding of the genetics of ASD.

Genomic Dissection of Bipolar Disorder and Schizophrenia, Including 28 Subphenotypes

Schizophrenia and bipolar disorder are two distinct diagnoses that share symptomology. Understanding the genetic factors contributing to the shared and disorder-specific symptoms will be crucial for improving diagnosis and treatment. In genetic data consisting of 53,555 cases (20,129 bipolar disorder [BD], 33,426 schizophrenia [SCZ]) and 54,065 controls, we identified 114 genome-wide significant loci implicating synaptic and neuronal pathways shared between disorders. Comparing SCZ to BD (23,585 SCZ, 15,270 BD) identified four genomic regions including one with disorder-independent causal variants and potassium ion response genes as contributing to differences in biology between the disorders. Polygenic risk score (PRS) analyses identified several significant correlations within case-only phenotypes including SCZ PRS with psychotic features and age of onset in BD. For the first time, we discover specific loci that distinguish between BD and SCZ and identify polygenic components underlying multiple symptom dimensions. These results point to the utility of genetics to inform symptomology and potential treatment.

Stability of ARDS subphenotypes over time in two randomised controlled trials

RATIONALE

Two distinct acute respiratory distress syndrome (ARDS) subphenotypes have been identified using data obtained at time of enrolment in clinical trials; it remains unknown if these subphenotypes are durable over time.

OBJECTIVE

To determine the stability of ARDS subphenotypes over time.

METHODS

Secondary analysis of data from two randomised controlled trials in ARDS, the ARMA trial of lung protective ventilation (n=473; patients randomised to low tidal volumes only) and the ALVEOLI trial of low versus high positive end-expiratory pressure (n=549). Latent class analysis (LCA) and latent transition analysis (LTA) were applied to data from day 0 and day 3, independent of clinical outcomes.

MEASUREMENTS AND MAIN RESULTS

In ALVEOLI, LCA indicated strong evidence of two ARDS latent classes at days 0 and 3; in ARMA, evidence of two classes was stronger at day 0 than at day 3. The clinical and biological features of these two classes were similar to those in our prior work and were largely stable over time, though class 2 demonstrated evidence of progressive organ failures by day 3, compared with class 1. In both LCA and LTA models, the majority of patients (>94%) stayed in the same class from day 0 to day 3. Clinical outcomes were statistically significantly worse in class 2 than class 1 and were more strongly associated with day 3 class assignment.

CONCLUSIONS

ARDS subphenotypes are largely stable over the first 3 days of enrolment in two ARDS Network trials, suggesting that subphenotype identification may be feasible in the context of clinical trials.

A Clustered Multiclass Likelihood-Ratio Ensemble Method for Family-Based Association Analysis Accounting for Phenotypic Heterogeneity

Although compelling evidence suggests that the genetic etiology of complex diseases could be heterogeneous in subphenotype groups, little attention has been paid to phenotypic heterogeneity in genetic association analysis of complex diseases. Simply ignoring phenotypic heterogeneity in association analysis could result in attenuated estimates of genetic effects and low power of association tests if subphenotypes with similar clinical manifestations have heterogeneous underlying genetic etiologies. To facilitate the family-based association analysis allowing for phenotypic heterogeneity, we propose a clustered multiclass likelihood-ratio ensemble (CMLRE) method. The proposed method provides an alternative way to model the complex relationship between disease outcomes and genetic variants. It allows for heterogeneous genetic causes of disease subphenotypes and can be applied to various pedigree structures. Through simulations, we found CMLRE outperformed the commonly adopted strategies in a variety of underlying disease scenarios. We further applied CMLRE to a family-based dataset from the International Consortium to Identify Genes and Interactions Controlling Oral Clefts (ICOC) to investigate the genetic variants and interactions predisposing to subphenotypes of oral clefts. The analysis suggested that two subphenotypes, nonsyndromic cleft lip without palate (CL) and cleft lip with palate (CLP), shared similar genetic etiologies, while cleft palate only (CP) had its own genetic mechanism. The analysis further revealed that rs10863790 (IRF6), rs7017252 (8q24), and rs7078160 (VAX1) were jointly associated with CL/CLP, while rs7969932 (TBK1), rs227731 (17q22), and rs2141765 (TBK1) jointly contributed to CP.