Faculty Bibliography

PURPOSE/OBJECTIVE:Biomarkers for disease specific survival (DSS) in early stage melanoma are needed to select patients for adjuvant immunotherapy and accelerate clinical trial design. We present a pathology-based computational method using a deep neural network architecture for DSS prediction. EXPERIMENTAL DESIGN/METHODS:The model was trained on 108 patients from four institutions and tested on 104 patients from Yale School of Medicine (YSM). A receiver operating characteristic (ROC) curve was generated based on vote aggregation of individual image sequences, an optimized cutoff was selected, and the computational model was tested on a third independent population of 51 patients from Geisinger Health Systems (GHS). RESULTS:Area under the curve (AUC) in the YSM patients was 0.905 (p<0.0001). AUC in the GHS patients was 0.880 (p<0.0001). Using the cutoff selected in the YSM cohort, the computational model predicted DSS in the GHS cohort based on Kaplan-Meier (KM) analysis (p<0.0001). CONCLUSIONS:The novel method presented is applicable to digital images, obviating the need for sample shipment and manipulation and representing a practical advance over current genetic and IHC-based methods.

OBJECTIVE:The primary somatosensory cortex (S1) and the anterior cingulate cortex (ACC) are two most important brain regions encoding the sensory-discriminative and affective-emotional aspects of pain, respectively. However, the functional connectivity of these two areas during cortical pain processing remains unclear. Developing methods to dissect the functional connectivity and directed information flow between cortical pain circuits can reveal insight into neural mechanisms of pain perception. APPROACH/METHODS:We recorded multichannel local field potentials (LFPs) from the S1 and ACC from freely behaving rats under various conditions of pain stimulus (thermal vs. mechanical) and pain state (naive vs. chronic pain). We applied Granger causality (GC) analysis to the LFP recordings and inferred frequency-dependent GC statistics and directed information flow. MAIN RESULTS/RESULTS:We found increased information flow during noxious pain stimulus presentation in both S1-->ACC and ACC-->S1 directions, especially at theta and gamma frequency bands. Similar results were found between thermal and mechanical pain stimuli. The chronic pain state shares common observations, but with elevated GC statistics especially in the gamma band. Furthermore, time-varying GC analysis revealed negative correlation between direction-specific and frequency-dependent GC and animal's paw withdrawal latency. In addition, we used computer simulations to investigate the impact of model mismatch, noise, missing variables, and common input on the conditional GC estimate. We also compared the GC results with the transfer entropy (TE) estimates. SIGNIFICANCE/CONCLUSIONS:Our results reveal functional connectivity and directed information flow between the S1 and ACC during various pain conditions. The time-varying GC analysis support the cortico-cortical information loop consistent with the computational predictive coding paradigm.

Pain is known to disrupt sleep patterns, and disturbances in sleep can further worsen pain symptoms. Sleep spindles occur during slow wave sleep and have established effects on sensory and affective processing in mammals. A number of chronic neuropsychiatric conditions, meanwhile, are known to alter sleep spindle density. The effect of persistent pain on sleep spindle waves, however, remains unknown, and studies of sleep spindles are challenging due to long period of monitoring and data analysis. Utilizing automated sleep spindle detection algorithms built on deep learning, we can monitor the effect of pain states on sleep spindle activity. In this study, we show that in a chronic pain model in rodents, there is a significant decrease in sleep spindle activity compared to controls. Meanwhile, methods to restore sleep spindles are associated with decreased pain symptoms. These results suggest that sleep spindle density correlates with chronic pain and may be both a potential biomarker for chronic pain and a target for neuromodulaton therapy.

OBJECTIVES/OBJECTIVE:To compare postoperative outcomes in patients prescribed long-acting opioids vs opioid-naïve patients who underwent elective noncardiac surgeries. DESIGN/METHODS:Retrospective cohort study. SETTING/METHODS:Single urban academic institution. METHODS AND SUBJECTS/METHODS:We retrospectively compared postoperative outcomes in long-acting opioid users vs opioid-naïve patients who underwent elective noncardiac surgeries. Inpatient and ambulatory surgery cohorts were separately analyzed. Preoperative medication lists were queried for the presence of long-acting opioids or absence of opioids. Multivariable logistic regression was performed to analyze the impact of long-acting opioid use on readmission rate, respiratory failure, and adverse cardiac events. Multivariable zero-truncated negative binomial regression was used to examine length of stay. RESULTS:After exclusions, there were 93,644 adult patients in the study population, 23,605 of whom underwent inpatient surgeries and 70,039 of whom underwent ambulatory surgeries. After adjusting for potential confounders and inpatient surgeries, preoperative long-acting opioid use was associated with increased risk of prolonged length of stay (incidence rate ratio = 1.1, 99% confidence interval [CI] = 1.0-1.2, P < 0.01) but not readmission. For ambulatory surgeries, preoperative long-acting opioid use was associated with increased risk of all-cause as well as pain-related readmission (odds ratio [OR] = 2.1, 99% CI = 1.5-2.9, P < 0.001; OR = 2.0, 99% CI = 0.85-4.2, P = 0.02, respectively). There were no significant differences for respiratory failure or adverse cardiac events. CONCLUSIONS:The use of preoperative long-acting opioids was associated with prolonged length of stay for inpatient surgeries and increased risk of all-cause and pain-related readmission for ambulatory surgeries. Timely interventions for patients on preoperative long-acting opioids may be needed to improve these outcomes.

Acute pain has an evolutionary role for the detection of and response to physical harm. In some cases, however, acute pain can impair function and lead to other morbidities. Chronic pain, meanwhile, can present as a psychopathological condition that significantly interferes with daily living. Most basic and translational pain research has focused on the molecular and cellular mechanisms in the spinal and peripheral nervous systems. In contrast, the brain plays a key role in the affective manifestation and cognitive control of pain. In particular, several cortical regions, such as the somatosensory cortex, prefrontal cortex, insular, and anterior cingulate cortex, are well-known to be activated by acute pain signals, and neurons in these regions have been demonstrated to undergo changes in response to chronic pain. Furthermore, these cortical regions can project to a number of forebrain and limbic structures to exert powerful top-down control of not only sensory pain transmission but also affective pain expression, and such cortical regulatory mechanisms are particularly relevant in chronic pain states. Newer techniques have emerged that allow detailed studies of central pain circuits in animal models, as well as how such circuits are modified by the presence of chronic pain and other predisposing psychosomatic factors. These mechanistic approaches can complement imaging in human studies. At the therapeutic level, a number of pharmacological and non-pharmacological interventions have recently been shown to engage these top-down control systems to provide analgesia. In this review, we will discuss how pain signals reach important cortical regions, and how these regions in turn project to sub-cortical areas of the brain to exert profound modulation of the pain experience. In addition, we will discuss the clinical relevance of such top-down pain regulation mechanisms.

We characterized the landscape and drug sensitivity of ERBB2 (HER2) mutations in cancers. In 11 datasets (n = 211,726), ERBB2 mutational hotspots varied across 25 tumor types. Common HER2 mutants yielded differential sensitivities to eleven EGFR/HER2 tyrosine kinase inhibitors (TKIs) in vitro, and molecular dynamics simulations revealed that mutants with a reduced drug-binding pocket volume were associated with decreased affinity for larger TKIs. Overall, poziotinib was the most potent HER2 mutant-selective TKI tested. Phase II clinical testing in ERBB2 exon 20-mutant non-small cell lung cancer resulted in a confirmed objective response rate of 42% in the first 12 evaluable patients. In pre-clinical models, poziotinib upregulated HER2 cell-surface expression and potentiated the activity of T-DM1, resulting in complete tumor regression with combination treatment.

Pain is a complex multidimensional experience, and pain perception is still incompletely understood. Here we combine animal behavior, electrophysiology, and computer modeling to dissect mechanisms of evoked and spontaneous pain. We record the local field potentials (LFPs) from the primary somatosensory cortex (S1) and anterior cingulate cortex (ACC) of freely behaving rats during pain episodes, and develop a predictive coding model to investigate the temporal coordination of oscillatory activity between the S1 and ACC. Our preliminary results from computational simulations support the experimental findings and provide new predictions.

Effective pharmacological treatment options for chronic pain remain very limited, and continued reliance on opioid analgesics has contributed to an epidemic in the U.S. On the other hand, non-pharmacologic neuromodulatory interventions provide a promising avenue for relief of chronic pain without the complications of dependence and addiction. An especially attractive neuromodulation strategy is to optimize endogenous pain regulatory circuits. The prefrontal cortex (PFC) is known to provide top-down control of pain, and hence neuromodulation methods that selectively enhance the activities in this brain region during pain episodes have the potential to provide analgesia. In this study, we designed a low-frequency (2 Hz) electrical stimulation protocol to provide temporally and spatially specific enhancement of the prefrontal control of pain in rats. We showed that low-frequency electrical stimulation of the prelimbic region of the PFC relieved both sensory and affective responses to acute pain in naïve rats. Furthermore, we found that low-frequency electrical stimulation of the PFC also attenuated mechanical allodynia in a rat model of chronic pain. Together, our findings demonstrated that low-frequency electrical stimulation of the PFC represents a promising new method of neuromodulation to inhibit pain.

Pain is a salient sensory experience with affective and cognitive dimensions. However, central mechanisms for pain remain poorly understood, hindering the development of effective therapeutics. Intracranial pharmacology presents an important tool for understanding the molecular and cellular mechanisms of pain in the brain, as well as for novel treatments. Here we present a protocol that integrates intracranial pharmacology with pain behavior testing. Specifically, we show how to infuse analgesic drugs into a select brain region, which may be responsible for pain modulation. Furthermore, to determine the effect of the candidate drug in the central nerve system, pain assays are performed after intracranial treatment. Our results demonstrate that intracranial administration of analgesic drugs in a targeted region can provide relief of pain in rodents. Thus, our protocol successfully demonstrates that intracranial pharmacology, combined with pain behavior testing, can be a powerful tool for the study of pain mechanisms in the brain.

OBJECTIVE:Sleep spindles have been implicated in memory consolidation and synaptic plasticity during NREM sleep. Detection accuracy and latency in automatic spindle detection are critical for real-time applications. APPROACH/METHODS:Here we propose a novel deep learning strategy (SpindleNet) to detect sleep spindles based on a single EEG channel. While the majority of spindle detection methods are used for off-line applications, our method is well suited for online applications. MAIN RESULTS/RESULTS:Compared with other spindle detection methods, SpindleNet achieves superior detection accuracy and speed, as demonstrated in two publicly available expert-validated EEG sleep spindle datasets. Our real-time detection of spindle onset achieves detection latencies of 150-350 ms (~2-3 spindle cycles) and retains excellent performance under low EEG sampling frequencies and low signal-to-noise ratios. SpindleNet has good generalization across different sleep datasets from various subject groups of different ages and species. SIGNIFICANCE/CONCLUSIONS:SpindleNet is ultra-fast and scalable to multichannel EEG recordings, with an accuracy level comparable to human experts, making it appealing for long-term sleep monitoring and closed-loop neuroscience experiments. &#13.