Biological Predictors Research Articles

Dopamine Transporter and CYP2D6 Gene Relationships with Attention-Deficit/Hyperactivity Disorder Treatment Response in the Methylphenidate and Atomoxetine Crossover Study.

Background: Few biological or clinical predictors guide medication selection and/or dosing for attention-deficit/hyperactivity disorder (ADHD). Accumulating data suggest that genetic factors may contribute to clinically relevant pharmacodynamic (e.g., dopamine transporter-SLC6A3 also commonly known as DAT1) or pharmacokinetic (e.g., the drug metabolizing enzyme Cytochrome P450 2D6 CYP2D6) effects of methylphenidate (stimulant) and atomoxetine (non-stimulant), which are commonly prescribed medications. This is the first study of youth with ADHD exposed to both medications examining the clinical relevance of genetic variation on treatment response. Methods: Genetic variations in DAT1 and CYP2D6 were examined to determine how they modified time relationships with changes in ADHD symptoms over a 4-week period in 199 youth participating in a double-blind crossover study following a stepped titration dose optimization protocol. Results: Our results identified trends in the modification effect from CYP2D6 phenotype and the time-response relationship between ADHD total symptoms for both medications (atomoxetine [ATX]: p = 0.058, Methylphenidate [MPH]: p = 0.044). There was also a trend for the DAT1 3' untranslated region (UTR) variable number of tandem repeat (VNTR) genotype to modify dose relationships with ADHD-RS total scores for atomoxetine (p = 0.029). Participants with DAT1 9/10 repeat genotypes had a more rapid dose-response to ATX compared to 10/10, while those with 9/9 genotypes did not respond as doses were increased. Regardless of genotype, ADHD symptoms and doses were similar across CYP2D6 metabolizer groups after 4 weeks of treatment. Conclusions: Most children with ADHD who were CYP2D6 normal metabolizers or had DAT1 10/10 or 9/10 genotypes responded well to both medications. While we observed some statistically significant effects of CYP2D6 and DAT1 with treatment response over time, our data indicate that genotyping for clinical purposes may have limited utility to guide treatment decisions for ATX or MPH because both medications were generally effective in the studied cohort after 3 weeks of titration to higher doses. The potential DAT1 association with ATX treatment is a novel finding, consistent with prior reports suggesting an association of the DAT1 in 9/9 genotypes with lower responsive rates to treatment at low and moderate doses.

Comprehensive Review of Future Liver Remnant (FLR) Assessment and Hypertrophy Techniques Before Major Hepatectomy: How to Assess and Manage the FLR.

The regenerative capacities of the liver and improvements in surgical techniques have expanded the possibilities of resectability. Liver resection is often the only curative treatment for primary and secondary malignancies, despite the risk of post-hepatectomy liver failure (PHLF). This serious complication (with a 50% mortality rate) can be avoided by better assessment of liver volume and function of the future liver remnant (FLR). The aim of this review was to understand and assess clinical, biological, and imaging predictors of PHLF risk, as well as the various hypertrophy techniques, to achieve an adequate FLR before hepatectomy. We reviewed the state of the art in liver regeneration and FLR hypertrophy techniques. The use of new biological scores (such as the aspartate aminotransferase/platelet ratio index + albumin-bilirubin [APRI+ALBI] score), concurrent utilization of 99mTc-mebrofenin scintigraphy (HBS), or dynamic hepatocyte contrast-enhanced MRI (DHCE-MRI) for liver volumetry helps predict the risk of PHLF. Besides portal vein embolization, there are other FLR optimization techniques that have their indications in case of risk of failure (e.g., associating liver partition and portal vein ligation for staged hepatectomy, liver venous deprivation) or in specific situations (transarterial radioembolization). There is a need to standardize volumetry and function measurement techniques, as well as FLR hypertrophy techniques, to limit the risk of PHLF.

A NONPARAMETRIC MIXED-EFFECTS MIXTURE MODEL FOR PATTERNS OF CLINICAL MEASUREMENTS ASSOCIATED WITH COVID-19.

Some patients with COVID-19 show changes in signs and symptoms such as temperature and oxygen saturation days before being positively tested for SARS-CoV-2, while others remain asymptomatic. It is important to identify these subgroups and to understand what biological and clinical predictors are related to these subgroups. This information will provide insights into how the immune system may respond differently to infection and can further be used to identify infected individuals. We propose a flexible nonparametric mixed-effects mixture model that identifies risk factors and classifies patients with biological changes. We model the latent probability of biological changes using a logistic regression model and trajectories in the latent groups using smoothing splines. We developed an EM algorithm to maximize the penalized likelihood for estimating all parameters and mean functions. We evaluate our methods by simulations and apply the proposed model to investigate changes in temperature in a cohort of COVID-19-infected hemodialysis patients.

Epigenetic age acceleration and mortality risk prediction in U.S. adults.

Epigenetic clocks have emerged as novel measures of biological age and potential predictors of mortality. We aimed to test whether epigenetic age acceleration (EAA) estimated using different epigenetic clocks predict long-term overall, cardiovascular or cancer mortality. We analyzed data from 2,105 participants to the 1999-2002 National Health and Nutrition Examination Survey aged ≥50 years old who were followed for mortality through 2019. EAAs was calculated from the residuals of Horvath, Hannum, SkinBlood, Pheno, Zhang, Lin, Weidner, Vidal-Bralo and Grim epigenetic clocks regressed on chronological age. Using cox proportional hazards regression, we estimated the hazard ratio (HR) and 95% confidence interval (CI) for the association of EAA (per 5-year) and the DunedinPoAm pace of aging (per 10% increase) with overall, cardiovascular and cancer mortality, adjusting for covariates and white blood cell composition. During a median follow-up of 17.5 years, 998 deaths occurred, including 272 from cardiovascular disease and 209 from cancer. Overall mortality was most significantly predicted by Grim EAA (P < 0.0001; HR: 1.50, 95% CI: 1.32-1.71) followed by Hannum (P = 0.001; HR: 1.16, 95% CI: 1.07-1.27), Pheno (P = 0.001; HR: 1.13, 95% CI: 1.05-1.21), Horvath (P = 0.007; HR: 1.13, 95% CI: 1.04-1.22) and Vidal-Bralo (P = 0.008; HR: 1.13, 95% CI: 1.03-1.23) EAAs. Grim EAA predicted cardiovascular mortality (P < 0.0001; HR: 1.55, 95% CI: 1.29-1.86), whereas Hannum (P = 0.006; HR: 1.24, 95% CI: 1.07-1.44), Horvath (P = 0.02; HR: 1.18, 95% CI: 1.02-1.35) and Grim (P = 0.049; HR: 1.37, 95% CI: 1.00-1.87) EAAs predicted cancer mortality. DunedinPoAm pace of aging was associated with overall (P = 0.003; HR: 1.23, 95% CI: 1.08-1.38) and cardiovascular (P = 0.04; HR: 1.25, 95% CI: 1.01-1.55) mortality. In a U.S. representative sample, Horvath, Hannum, Pheno, Vidal-Bralo and Grim EAA all predicted overall mortality but only Grim EAA predicted cardiovascular mortality and Horvath, Hannum or Grim EAA predicted cancer mortality. Pace of aging predicted overall and cardiovascular mortality.

The PRECISE-DYAD Neurodevelopmental substudy protocol: neurodevelopmental risk in children of mothers with pregnancy complications.

Over 250 million children are not reaching their developmental potential globally. The impact of prenatal factors and their interplay with postnatal environmental factors on child neurodevelopment, is still unclear-particularly in low- and middle-income settings. This study aims to understand the impact of pregnancy complications as well as environmental, psychosocial, and biological predictors on neurodevelopmental trajectories. This is an observational cohort study of female and male children (≈3,950) born to women (≈4,200) with and without pregnancy complications (pregnancy-induced hypertension, foetal growth restriction, and premature birth) previously recruited into PREgnancy Care Integrating Translational Science, Everywhere study with detailed biological data collected in intrapartum and post-partum periods. Children will be assessed at six weeks to 6 months, 11-13 months, 23-25 months and 35-37 months in rural and semi-urban Gambia (Farafenni, Illiasa, and Ngayen Sanjal) and Kenya (Mariakani and Rabai). We will assess children's neurodevelopment using Prechtls General Movement Assessment, the Malawi Development Assessment Tool (primary outcome), Observation of Maternal-Child Interaction, the Neurodevelopmental Disorder Screening Tool, and the Epilepsy Screening tool. Children screening positive will be assessed with Cardiff cards (vision), Modified Checklist for Autism in Toddlers Revised, and Pediatric Quality of Life Inventory Family Impact. We will use multivariate logistic regression analysis to investigate the impact of pregnancy complications on neurodevelopment and conduct structural equation modelling using latent class growth to study trajectories and relationships between biological, environmental, and psychosocial factors on child development. We aim to provide information regarding the neurodevelopment of infants and children born to women with and without pregnancy complications at multiple time points during the first three years of life in two low-resource African communities. A detailed evaluation of developmental trajectories and their predictors will provide information on the most strategic points of intervention to prevent and reduce the incidence of neurodevelopmental impairments.

Biological and Procedural Predictors of Outcome in the Stroke Preclinical Assessment Network (SPAN) Trial.

The SPAN trial (Stroke Preclinical Assessment Network) is the largest preclinical study testing acute stroke interventions in experimental focal cerebral ischemia using endovascular filament middle cerebral artery occlusion (MCAo). Besides testing interventions against controls, the prospective design captured numerous biological and procedural variables, highlighting the enormous heterogeneity introduced by the multicenter structure that might influence stroke outcomes. Here, we leveraged the unprecedented sample size achieved by the SPAN trial and the prospective design to identify the biological and procedural variables that affect experimental stroke outcomes in transient endovascular filament MCAo. The study cohort included all mice enrolled and randomized in the SPAN trial (N=1789). Mice were subjected to 60-minute MCAo and followed for a month. Thirteen biological and procedural independent variables and 4 functional (weight loss and 4-point neuroscore on days 1 and 2, corner test on days 7 and 28, and mortality) and 3 tissue (day 2, magnetic resonance imaging infarct volumes and swelling; day 30, magnetic resonance imaging tissue loss) outcome variables were prospectively captured. Multivariable regression with stepwise elimination was used to identify the predictors and their effect sizes. Older age, active circadian stage at MCAo, and thinner and longer filament silicone tips predicted higher mortality. Older age, larger body weight, longer anesthesia duration, and longer filament tips predicted worse neuroscores, while high-fat diet and blood flow monitoring predicted milder neuroscores. Older age and a high-fat diet predicted worse corner test performance. While shorter filament tips predicted more ipsiversive turning, longer filament tips appeared to predict contraversive turning. Age, sex, and weight interacted when predicting the infarct volume. Older age was associated with smaller infarcts on day 2 magnetic resonance imaging, especially in animals with larger body weights; this association was most conspicuous in females. High-fat diet also predicted smaller infarcts. In contrast, the use of cerebral blood flow monitoring and more severe cerebral blood flow drop during MCAo, longer anesthesia, and longer filament tips all predicted larger infarcts. Bivariate analyses among the dependent variables highlighted a disconnect between tissue and functional outcomes. Our analyses identified variables affecting endovascular filament MCAo outcome, an experimental stroke model used worldwide. Multiple regression refuted some commonly reported predictors and revealed previously unrecognized associations. Given the multicenter prospective design that represents a sampling of real-world conditions, the degree of heterogeneity mimicking clinical trials, the large number of predictors adjusted for in the multivariable model, and the large sample size, we think this is the most definitive analysis of the predictors of preclinical stroke outcome to date. Future multicenter experimental stroke trials should standardize or at least ensure a balanced representation of the biological and procedural variables identified herein as potential confounders.

Co-analysis of methylation platforms for signatures of biological aging in the domestic dog reveals previously unexplored confounding factors.

Chronological age reveals the number of years an individual has lived since birth. By contrast, biological age varies between individuals of the same chronological age at a rate reflective of physiological decline. Differing rates of physiological decline are related to longevity and result from genetics, environment, behavior, and disease. The creation of methylation biological age predictors is a long-standing challenge in aging research due to the lack of individual pre-mortem longevity data. The consistent differences in longevity between domestic dog breeds enable the construction of biological age estimators which can, in turn, be contrasted with methylation measurements to elucidate mechanisms of biological aging. We draw on three flagship methylation studies using distinct measurement platforms and tissues to assess the feasibility of creating biological age methylation clocks in the dog. We expand epigenetic clock building strategies to accommodate phylogenetic relationships between individuals, thus controlling for the use of breed standard metrics. We observe that biological age methylation clocks are affected by population stratification and require heavy parameterization to achieve effective predictions. Finally, we observe that methylation-related markers reflecting biological age signals are rare and do not colocalize between datasets.

Monitoring immunE DysregulAtion foLLowing Immune checkpOint-inhibitioN (MEDALLION): protocol for an observational cancer immunotherapy cohort study

BackgroundCheckpoint inhibitors (CPIs) are widely used in cancer treatment, with transformative impacts on survival. They nonetheless carry a significant risk of toxicity in the form of immune-related adverse events (IrAEs), which may be sustained and life-altering. IrAEs may require high-dose and/or prolonged steroid use and represent a significant healthcare burden. They mimic immune-mediated inflammatory diseases (IMIDs) but understanding of their pathogenesis is limited. The MEDALLION project aims to determine targetable mechanisms of immune dysregulation in IrAE development, employing an immune monitoring approach to determine changes in circulating and tissue resident cells of CPI recipients who do/do not develop them and assessing the contribution of the microbiome in parallel.MethodsMEDALLION is a non-randomised longitudinal cohort study aiming to recruit 66 cancer patient recipients of anti-PD1/PD-L1, anti-CTLA-4 or combination therapy. Eligible participants include those with malignant melanoma in the adjuvant or metastatic setting, mesothelioma and non-small cell lung carcinoma (NSCLC) treated in the metastatic setting. Comprehensive clinical evaluation is carried out alongside blood, skin swab and stool sampling at the time of CPI initiation (baseline) and during subsequent routine hospital visits on 6 occasions over a 10-month follow-up period. It is conservatively anticipated that one third of enrolled patients will experience a “significant IrAE” (SirAE), defined according to pre-determined criteria specific to the affected tissue/organ system. Those developing such toxicity may optionally undergo a biopsy of affected tissue where appropriate, otherwise being managed according to standard of care. Peripheral blood mononuclear cells will be analysed using multi-parameter flow cytometry to investigate immune subsets, their activation status and cytokine profiles. Stool samples and skin swabs will undergo DNA extraction for 16 S ribosomal RNA (rRNA) sequencing and internal transcribed spacer (ITS) gene sequencing to determine bacterial and fungal microbiome diversity, respectively, including species associated with toxicity. Stored tissue biopsies will be available for in situ and single-cell transcriptomic evaluation. Analysis will focus on the identification of biological predictors and precursors of SirAEs.DiscussionThe pathogenesis of IrAEs will be assessed through the MEDALLION cohort, with the potential to develop tools for their prediction and/or strategies for targeted prevention or treatment.Trial RegistrationThe study was registered on 18/09/2023 in the ISRCTN registry (43,419,676).

Association between single and mixed exposure to polycyclic aromatic hydrocarbons and biological aging.

Aging is one of the most important public health issues. Previous studies on the factors affecting aging focused on genetics and lifestyle, but the association between polycyclic aromatic hydrocarbons (PAHs) and aging is still unclear. This study utilized data from the National Health and Nutrition Examination Survey (NHANES) 2003-2010. A total of 8,100 participants was used to construct the biological age predictors by using recent advanced algorithms Klemera-Doubal method (KDM) and Mahalanobis distance. Two biological aging indexes, recorded as KDM-BA acceleration and PhenoAge acceleration, were used to investigate the relationship between single PAHs and biological age using a multiple linear regression analysis, and a weighted quantile sum (WQS) model was constructed to explore the mixed effects of PAHs on biological age. Finally, we constructed the restricted cubic spline (RCS) model to assess the non-linear relationship between PAHs and biological age. Exposure to PAHs was associated with PhenoAge acceleration. Each unit increase in the log10-transformed level of 1-naphthol, 2-naphthol, and 2-fluorene was associated with a 0.173 (95% CI: 0.085, 0.261), 0.310 (95% CI: 0.182, 0.438), and 0.454 (95% CI: 0.309, 0.598) -year increase in PhenoAge acceleration, respectively (all corrected P < 0.05). The urinary PAH mixture was relevant to KDM-BA acceleration (β = 0.13, 95% CI: 0, 0.26, P = 0.048) and PhenoAge acceleration (β = 0.59, 95% CI: 0.47, 0.70, P < 0.001), and 2-naphthol had the highest weight in the weighted quantile sum (WQS) regression. The RCS analyses showed a non-linear association between 2-naphthol and 2-fluorene with KDM-BA acceleration (all P < 0.05) in addition to a non-linear association between 1-naphthol, 2-naphthol, 3-fluorene, 2-fluorene, and 1-pyrene with PhenoAge acceleration (all P < 0.05). Exposure to mixed PAHs is associated with increased aging, with 2-naphthol being a key component of PAHs associated with aging. This study has identified risk factors in terms of PAH components for aging.

WED-223 Clinical and biological predictors of liver fat content ≥8% as assessed by MRI-PDFF: combined data from multiple therapeutic trials including more than 10, 000 patients

WED-223 Clinical and biological predictors of liver fat content ≥8% as assessed by MRI-PDFF: combined data from multiple therapeutic trials including more than 10, 000 patients

Clinical and biological predictors of Cladribine effectiveness in Multiple Sclerosis: A real-world, single Centre study considering a two-year interval from year-2 dosing

ObjectivesCladribine tablets (CLAD) for adult patients with highly active relapsing multiple sclerosis (RMS) have been available in Italy since 2018. We aimed to assess predictors of no-evidence-of-disease-activity-3 (NEDA-3) status after 24 months of the last dose of CLAD. ResultsWe included 88 patients (70.5% female, mean age at CLAD start 35.4 ± 11.4). Eighteen patients were treatment naïve, 48 switched to CLAD from a First line Disease Modifying Drug (DMD), and 22 from Second line DMDs. All patients were observed for a median follow-up time of 2.4 (1–4) years after the last dose of CLAD. Forty-nine patients (55.7%) showed NEDA at the last available follow-up. Naïve patients (p = 0.001), those with a lower number of previous DMDs (p < 0.001) and, even though not significantly, those switching from first line DMDs (p = 0.069) were more likely NEDA3 at the last available follow-up. In a subgroup of 30 patients (34%), Serum Light Neurofilaments (sNFL) levels showed a decrease from baseline to the 24 months of follow-up, statistically significant from baseline to the sixth month, and from the first to the second year detection. sNFL levels at 12th month showed a strong inverse correlation with the time to NEDA3 loss. ConclusionsOur experience provides information for the 2-years after the last dose of CLAD, confirming a higher effectiveness of CLAD when placed early in the treatment algorithm. Given the ongoing expansion of the therapeutic landscape in MS, sNfL could support individualized decision-making, used as blood-based biomarker for CLAD responses in clinical practice.

Abstract PO2-09-08: Biological and non-biological predictors of breast cancer disease-free survival among young Black females

Abstract Introduction: Young Black females bear a disproportionate burden of breast cancer (BC) deaths compared to White females yet remain underrepresented in clinical studies. We sought to discover predictors of disease-free survival (DFS) in a cohort of young Black females with BC. Methods: Black females diagnosed with invasive BC ≤ age 50 from 2005 to 2016 were recruited through the Florida and Tennessee state cancer registries. Participants were asked to complete a questionnaire, medical records release, and tissue/tumor release. Saliva was also requested for germline DNA extraction. For the primary outcome of BC DFS, univariate analyses were conducted using the following variables: proportion of West African ancestry, RNA expression based on PAM50 analyses, body mass index (BMI), social determinants of health (i.e., employment and insurance), immunohistochemistry (IHC) subtype, treatment category (i.e., chemotherapy, radiation, and surgery), and BC family history. Multivariate analyses were conducted using backward selection among these variables to analyze in the reduced model. Results: Of 701 consented, 687 participants with early-stage disease were included in the analysis; 14 participants with stage 4 disease at diagnosis were excluded. Among the 687 participants, the median age at diagnosis was 44, 28% had triple-negative breast cancer (TNBC), 13% were deceased, and 5% had recurrent disease. Clinically, participants with TNBC and lymph node involvement had worse BC DFS (p=0.001 and p=0.0002, respectively). Socio-economically, full-time employement was associated with higher BC DFS (p=0.0003). Among those for whom global ancestry data were available (n=551), multivariate analysis showed an association between increased percent of West African ancestry and worse BC DFS with HRIQR=1.23*, which approached significance (p=0.085). Additional subgroup analyses among those with hormone receptor-positive (HR+) (estrogen and/or progesterone receptor positive) BC (n=431) showed that participants with private insurance had better BC DFS (HR=0.45; p=0.05), while those with lymph node involvement had worse BC DFS (HR=2.07; p=0.001). Among HR+ participants with available ancestry data (n=349), worse BC DFS was associated with increased West African ancestry (p=0.05) and lymph node involvement (p=0.08). When ancestry was included in the model, private insurance was no longer associated with BC DFS in this HR+ subgroup (p=0.31). Other predictors analyzed did not reach statistical significance. Conclusion: Our findings confirm prior established adverse predictors of BC DFS, such as TNBC with lymph node involvement and the protective effect of full-time employment. We also found a novel association with West African ancestry among patients with HR+ breast cancer. Although further large-scale studies are needed, results from this cohort of young Black females with BC and highlight the critical need to support research to understand biological and non-biological factors contributing to BC DFS and develop targeted strategies to improve survival outcomes. * HRIQR is the ratio of hazard rates corresponding to upper and lower quartiles of percent West African (interquartile range (IQR): 70.0%-80.8%). Citation Format: Sonya Reid, Run Fan, Lindsay Venton, Anne Weidner, Ann Tezak, Mya Roberson, Susan Vadaparampil, Xuefeng Wang, Sean Yoder, Marilin Rosa, Jibril Hirbo, Jennifer Whisenant, Jennifer Pietenpol, Sheila Rajagopal, Brian Lehmann, Fei Ye, Tuya Pal. Biological and non-biological predictors of breast cancer disease-free survival among young Black females [abstract]. In: Proceedings of the 2023 San Antonio Breast Cancer Symposium; 2023 Dec 5-9; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2024;84(9 Suppl):Abstract nr PO2-09-08.

Abstract PS11-01: Brain metastases in metastatic breast cancer: prevalence per line of treatment and cumulative incidence in a cohort of 18075 real-world patients

Abstract BACKGROUND Historic data suggest that incidence of brain metastasis (BM) is relatively distributed over time in HER2-positive (HER2+) breast cancer (BC), occurs early in triple negative BC (TNBC), and occurs late in hormone receptor-positive (HR+) metastatic BC (mBC). However, the timing of BM relative to line of therapy has not been well described. We describe the prevalence per line of therapy and cumulative incidence of BM in a large real-world cohort of patients (pts) with mBC, by BC subtype and line of therapy. METHODS This analysis used data from the longitudinal US Flatiron Health de-identified database, which provides unstructured and structured electronic health record-derived data from &amp;gt;2.6 million pts with cancer in ~800 unique sites of care. Eligible pts had initiated a first line of treatment (index data) for mBC up to March 1, 2021, to allow for ≥2 years potential follow-up time. Baseline characteristics were assessed, and pts categorized by HER2 and HR status: 1) HR+, HER2+; 2) HR+, HER2–, 3) HR–, HER2+ and 4) TNBC (i.e., negative for HR and HER2). For the last two categories, HER2-low subsets were defined based on immunohistochemistry results (1+, 2+, or equivocal). Lines of therapy were derived using both treatment regimens and progression data. The primary outcome was the first diagnosis of BM. The prevalence of BM was evaluated by subtype, at the index date, and by the line of therapy. The cumulative incidence function (CIF) of BM was used to estimate the risk of BM in pts free of BM at the index date, and death was treated as a competing event. RESULTS Overall, 18075 pts were included (HR+, HER2+: 3062 pts [16.9%]; HR–, HER2+: 902 pts [5.0%]; HR+, HER2–: 12331 pts [68.2%] [HR+, HER2-low: 7062 (39.1%)]; TNBC: 1780 pts [9.8%] [HR–, HER2-low: 725 pts (4.0%)]). Median age at the index date was 64 years (interquartile range: 54, 73) and 10271pts (56.8%) had visceral metastasis. In total, 5951 pts (32.9%) had de novo disease, 12090 (66.9%) had recurrent disease, and 34 (0.2%) did not have available data for mBC type. The table shows the number of pts by line of therapy, prevalence of BM, and cumulative incidence of BM from the index date. Of the included pts, 1306 (7.2%) had a BM at the index date; the CIF was run on the remaining 16973 pts. Overall, 2248 pts (13.2%) had an incident BM event during follow-up (HR+, HER2+: 578 events; HR–, HER2+: 237 events; HR+, HER2–: 1119 events [HR+, HER2-low: 619 events]; TNBC: 314 events [HR–, HER2-low: 124 events]), 9314 had a competing event, and 5411 were censored. By fourth-line therapy, the prevalence of BM was 26.1% in HR+, HER2+; 37.1% in HR–, HER2+; 24.7% in TNBC (HR–, HER2-low: 27.9%); but remained low at 7.2% in HR+, HER2– (HR+, HER2-low: 9.4%). The cumulative incidence of BM at 60 months was 23% in HR+, HER2+, 34% in HR–, HER2+, 10% in HR+, HER2–, and, 22% in TNBC. Overall, the HER2-low subsets had BM prevalence and cumulative incidence that were very close to those that were HER2–. CONCLUSIONS This large analysis of real-world data demonstrates that the prevalence of BM across pts with mBC differs by tumor subtype and by line of therapy. Conversely, HER2-low status may have limited impact. The cumulative incidence of BM was highest in the HR–, HER2+ and TNBC subgroups, and lowest in the HR+, HER2– subgroup; despite this, due to the high proportion of HER2–, HR+ mBC cases, the HR+, HER2– group represented the largest number of BM events. These data emphasize the need for clinical and biologic predictors of BM and for strategies to prevent their onset, and provide information on the impact of BM inclusion and exclusion criteria in clinical trials for pts with mBC. Number of patients by line of therapy, prevalence of BM, and cumulative incidence of BM from the index date BM, brain metastasis; HR, hormone receptor, pts, patients; TNBC, triple negative breast cancer. Citation Format: Sarah Sammons, Jose Leone, Thibaut Sanglier, Peter Lambert, Filippo Montemurro, Raf Poppe, Eleonora Restuccia, Sara Tolaney, Nancy Lin. Brain metastases in metastatic breast cancer: prevalence per line of treatment and cumulative incidence in a cohort of 18075 real-world patients [abstract]. In: Proceedings of the 2023 San Antonio Breast Cancer Symposium; 2023 Dec 5-9; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2024;84(9 Suppl):Abstract nr PS11-01.

OMICmAge: An integrative multi-omics approach to quantify biological age with electronic medical records

Biological aging is a multifactorial process involving complex interactions of cellular and biochemical processes that is reflected in omic profiles. Using common clinical laboratory measures in ~30,000 individuals from the MGB-Biobank with balanced sex representation, we developed a robust, predictive biological aging phenotype, EMRAge, that balances clinical biomarkers with overall mortality risk and can be broadly recapitulated across EMRs. We then applied elastic-net regression to model EMRAge with DNA-methylation (DNAm) and multiple omics, generating DNAmEMRAge and OMICmAge, respectively. Both biomarkers demonstrated strong associations with chronic diseases and mortality that outperform current biomarkers across our discovery (MGB-ABC, n=3,451) and validation (TruDiagnostic, n=12,666) cohorts. Furthermore, both DNAmEMRAge and OMICmAge exhibit greater accuracy (average ROC = 0.88) in predicting 5-year and 10-year mortality than current biological age predictors. Through the use of epigenetic biomarker proxies, OMICmAge has the unique advantage of expanding the predictive search space to include epigenomic, proteomic, metabolomic, and clinical data while distilling this in a measure with DNAm alone, providing opportunities to identify clinically-relevant interconnections central to the aging process. This work was funded in part by TruDiagnostic for data generation and manuscript writing. Efforts for RK, KM, YC, SB, and JALS are supported by R01HL123915 from the NIH/NHLBI; PK is supported by K99HL159234 from the NIH/NHLBI; RSK is supported by K01HL146980 from the NIH/NHLBI; SHC is supported by K01HL153941 from the NIH/NHLBI; YC and JAILS are supported by R01HL141826 from the NIH/NHLBI; AD is supported by K01HL130629 from the NIH/NHLBI; AD and JALS are supported by 1R01HL152244 from the NIH/NHLBI; MM and JALS are supported by R01HL155742 from the NIH/NHLBI; MM is supported by R01HL139634 from the NIH/NHLBI; JALS, STW, EWK are supported by the NIH U01HG008685; VNG is supported by NIH AG065403 and AG064223; CEW is supported by the JSPS KAKENHI Grant (JP19K21239), the Japanese Environment Research and Technology Development Fund (No. 5- 1752), JPJ SBP-1201854), HLF 20180290, HLF 20200693), and the Swedish Research Council (2016-02798); NJWR is supported by MR/Y010736/1 and IF\R1\231034; EM is supported by the NIH (ZICTR000410-03). This is the full abstract presented at the American Physiology Summit 2024 meeting and is only available in HTML format. There are no additional versions or additional content available for this abstract. Physiology was not involved in the peer review process.

The AccelerAge framework: a new statistical approach to predict biological age based on time-to-event data

Aging is a multifaceted and intricate physiological process characterized by a gradual decline in functional capacity, leading to increased susceptibility to diseases and mortality. While chronological age serves as a strong risk factor for age-related health conditions, considerable heterogeneity exists in the aging trajectories of individuals, suggesting that biological age may provide a more nuanced understanding of the aging process. However, the concept of biological age lacks a clear operationalization, leading to the development of various biological age predictors without a solid statistical foundation. This paper addresses these limitations by proposing a comprehensive operationalization of biological age, introducing the “AccelerAge” framework for predicting biological age, and introducing previously underutilized evaluation measures for assessing the performance of biological age predictors. The AccelerAge framework, based on Accelerated Failure Time (AFT) models, directly models the effect of candidate predictors of aging on an individual’s survival time, aligning with the prevalent metaphor of aging as a clock. We compare predictors based on the AccelerAge framework to a predictor based on the GrimAge predictor, which is considered one of the best-performing biological age predictors, using simulated data as well as data from the UK Biobank and the Leiden Longevity Study. Our approach seeks to establish a robust statistical foundation for biological age clocks, enabling a more accurate and interpretable assessment of an individual’s aging status.

Clarifying the biological and statistical assumptions of cross-sectional biological age predictors: an elaborate illustration using synthetic and real data

BackgroundThere is divergence in the rate at which people age. The concept of biological age is postulated to capture this variability, and hence to better represent an individual’s true global physiological state than chronological age. Biological age predictors are often generated based on cross-sectional data, using biochemical or molecular markers as predictor variables. It is assumed that the difference between chronological and predicted biological age is informative of one’s chronological age-independent aging divergence ∆.MethodsWe investigated the statistical assumptions underlying the most popular cross-sectional biological age predictors, based on multiple linear regression, the Klemera-Doubal method or principal component analysis. We used synthetic and real data to illustrate the consequences if this assumption does not hold.ResultsThe most popular cross-sectional biological age predictors all use the same strong underlying assumption, namely that a candidate marker of aging’s association with chronological age is directly informative of its association with the aging rate ∆. We called this the identical-association assumption and proved that it is untestable in a cross-sectional setting. If this assumption does not hold, weights assigned to candidate markers of aging are uninformative, and no more signal may be captured than if markers would have been assigned weights at random.ConclusionsCross-sectional methods for predicting biological age commonly use the untestable identical-association assumption, which previous literature in the field had never explicitly acknowledged. These methods have inherent limitations and may provide uninformative results, highlighting the importance of researchers exercising caution in the development and interpretation of cross-sectional biological age predictors.

Confounders in Predictive Medical Models: Roles of Nationality and Immigrant Status

Aim: The aim of this study was to assess the opinion of natural science specialists on the latest recommendations of official regulators regarding the prevention of causes of social disparities in artificial intelligence (AI) and machine learning (ML) models. Materials and Methods: An anonymous online survey was conducted using the Telegram platform, where participants were asked a single question: "Is the inclusion of predictors such as “nationality” and “immigrant status” in AI and ML medical models ethical and consistent with contemporary scientific standards?" Respondents were provided with two response options: "Yes" or "No." The survey was specifically targeted at international groups, focusing primarily on English and Russian-speaking clinicians and scientific researchers. Results: 180 unique individuals participated in the survey. The results revealed that one-third of the respondents (60 individuals) agreed that including predictors such as nationality and immigration status is inappropriate in the current ML and AI models. Conclusion: In conclusion, the fact that only one-third of respondents disagree with categorizing patients based on nationality is at odds with the standards set by official regulators. This discrepancy underscores the need for educational programs aimed at sensitizing the scientific community to prioritize biological predictors over data documented in passports or identity cards.

Incidence and predictors of metabolic syndrome onset in individuals with bipolar disorders: A longitudinal study from the FACE-BD cohort.

Metabolic syndrome (MetS) is a cluster of components including abdominal obesity, hyperglycemia, hypertension, and dyslipidemia. MetS is highly prevalent in individuals with bipolar disorders (BD) with an estimated global rate of 32.6%. Longitudinal data on incident MetS in BD are scarce and based on small sample size. The objectives of this study were to estimate the incidence of MetS in a large longitudinal cohort of 1521 individuals with BD and to identify clinical and biological predictors of incident MetS. Participants were recruited from the FondaMental Advanced Center of Expertise for Bipolar Disorder (FACE-BD) cohort and followed-up for 3 years. MetS was defined according to the International Diabetes Federation criteria. Individuals without MetS at baseline but with MetS during follow-up were considered as having incident MetS. A logistic regression model was performed to estimate the adjusted odds ratio and its corresponding 95% confidence interval (CI) for an association between each factor and incident MetS during follow-up. We applied inverse probability-of-censoring weighting method to minimize selection bias due to loss during follow-up. Among individuals without MetS at baseline (n = 1521), 19.3% developed MetS during follow-up. Multivariable analyses showed that incident MetS during follow-up was significantly associated with male sex (OR = 2.2, 95% CI = 1.7-3.0, p < 0.0001), older age (OR = 2.14, 95% CI = 1.40-3.25, p = 0.0004), presence of a mood recurrence during follow-up (OR = 1.91, 95% CI = 1.22-3.00, p = 0.0049), prolonged exposure to second-generation antipsychotics (OR = 1.56, 95% CI = 0.99, 2.45, p = 0.0534), smoking status at baseline (OR = 1.30, 95% CI = 1.00-1.68), lifetime alcohol use disorders (OR = 1.33, 95% CI = 0.98-1.79), and baseline sleep disturbances (OR = 1.04, 95% CI = 1.00-1.08), independently of the associations observed for baseline MetS components. We observed a high incidence of MetS during a 3 years follow-up (19.3%) in individuals with BD. Identification of predictive factors should help the development of early interventions to prevent or treat early MetS.

High dimensional predictions of suicide risk in 4.2 million US Veterans using ensemble transfer learning

We present an ensemble transfer learning method to predict suicide from Veterans Affairs (VA) electronic medical records (EMR). A diverse set of base models was trained to predict a binary outcome constructed from reported suicide, suicide attempt, and overdose diagnoses with varying choices of study design and prediction methodology. Each model used twenty cross-sectional and 190 longitudinal variables observed in eight time intervals covering 7.5 years prior to the time of prediction. Ensembles of seven base models were created and fine-tuned with ten variables expected to change with study design and outcome definition in order to predict suicide and combined outcome in a prospective cohort. The ensemble models achieved c-statistics of 0.73 on 2-year suicide risk and 0.83 on the combined outcome when predicting on a prospective cohort of ∼\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\\sim$$\\end{document} 4.2 M veterans. The ensembles rely on nonlinear base models trained using a matched retrospective nested case-control (Rcc) study cohort and show good calibration across a diversity of subgroups, including risk strata, age, sex, race, and level of healthcare utilization. In addition, a linear Rcc base model provided a rich set of biological predictors, including indicators of suicide, substance use disorder, mental health diagnoses and treatments, hypoxia and vascular damage, and demographics.

Biological and psychological predictors of cognitive function in breast cancer patients before surgery.

Research suggests that cancer-related cognitive impairment (CRCI) can occur before breast cancer (BC) treatment. The limited extant evidence suggests the underlying mechanisms could be stress-related. Potential psychological and biological predictors of CRCI prior to any BC treatment were examined. 112 treatment-naïve women with BC and 67 healthy controls (HC) completed a neuropsychological test battery to assess cognitive impairment and a self-report battery to assess cognitive complaints, cancer-related stress, depressive and anxiety symptoms. Morning and evening cortisol and α-amylase were collected from saliva. Multilinear regressions were conducted. Treatment-naïve BC patients were more frequently impaired in verbal memory and processing speed and reported more cognitive complaints (all p < .001) than HC. BC patients and HC did not differ in overall cognitive impairment (p = .21). Steeper α-amylase, lower cancer-related stress and younger age was associated with better overall cognitive function in treatment-naïve BC patients. Higher depressive symptoms predicted higher levels of cognitive complaints in BC patients. Overall, these findings suggest that stress plays a role in CRCI. This study is the first to associate α-amylase with cognitive function in cancer patients, informing future research. The findings on impairment in processing speed and verbal memory among treatment-naïve BC highlight the need to screen for such impairments among BC patients and indicate that future studies on CRCI should include baseline assessments prior to BC treatment. If replicated, these findings could inform the development and testing of appropriate interventions to decrease CRCI among cancer patients. NCT04418856, date of registration: 06.05.2020.