Identification of acute respiratory distress syndrome subphenotypes de novo using routine clinical data: a retrospective analysis of ARDS clinical trials.

OBJECTIVES: The acute respiratory distress syndrome (ARDS) is a heterogeneous condition, and identification of subphenotypes may help in better risk stratification. Our study objective is to identify ARDS subphenotypes using new simpler methodology and readily available clinical variables. SETTING: This is a retrospective Cohort Study of ARDS trials. Data from the US ARDSNet trials and from the international ART trial. PARTICIPANTS: 3763 patients from ARDSNet data sets and 1010 patients from the ART data set. PRIMARY AND SECONDARY OUTCOME MEASURES: The primary outcome was 60-day or 28-day mortality, depending on what was reported in the original trial. K-means cluster analysis was performed to identify subgroups. Sets of candidate variables were tested to assess their ability to produce different probabilities for mortality in each cluster. Clusters were compared with biomarker data, allowing identification of subphenotypes. RESULTS: Data from 4773 patients were analysed. Two subphenotypes (A and B) resulted in optimal separation in the final model, which included nine routinely collected clinical variables, namely heart rate, mean arterial pressure, respiratory rate, bilirubin, bicarbonate, creatinine, PaO2, arterial pH and FiO2. Participants in subphenotype B showed increased levels of proinflammatory markers, had consistently higher mortality, lower number of ventilator-free days at day 28 and longer duration of ventilation compared with patients in the subphenotype A. CONCLUSIONS: Routinely available clinical data can successfully identify two distinct subphenotypes in adult ARDS patients. This work may facilitate implementation of precision therapy in ARDS clinical trials.

Gene Expression Scoring of Immune Activity Levels for Precision Use of Hydrocortisone in Vasodilatory Shock.

PURPOSE: Among patients with vasodilatory shock, gene expression scores may identify different immune states. We aimed to test whether such scores are robust in identifying patients' immune state and predicting response to hydrocortisone treatment in vasodilatory shock. MATERIALS AND METHODS: We selected genes to generate continuous scores to define previously established subclasses of sepsis. We used these scores to identify a patient's immune state. We evaluated the potential for these states to assess the differential effect of hydrocortisone in two randomized clinical trials of hydrocortisone versus placebo in vasodilatory shock. RESULTS: We initially identified genes associated with immune-adaptive, immune-innate, immune-coagulant functions. From these genes, 15 were most relevant to generate expression scores related to each of the functions. These scores were used to identify patients as immune-adaptive prevalent (IA-P) and immune-innate prevalent (IN-P). In IA-P patients, hydrocortisone therapy increased 28-day mortality in both trials (43.3% vs 14.7%, P = 0.028) and (57.1% vs 0.0%, P = 0.99). In IN-P patients, this effect was numerically reversed. CONCLUSIONS: Gene expression scores identified the immune state of vasodilatory shock patients, one of which (IA-P) identified those who may be harmed by hydrocortisone. Gene expression scores may help advance the field of personalized medicine.

Accurate identification of breast cancer margins in microenvironments of ex-vivo basal and luminal breast cancer tissues using Raman spectroscopy.

Better knowledge of the breast tumor microenvironment is required for surgical resection and understanding the processes of tumor development. Raman spectroscopy is a promising tool that can assist in uncovering the molecular basis of disease and provide quantifiable molecular information for diagnosis and treatment evaluation. In this work, eighty-eight frozen breast tissue sections, including forty-four normal and forty-four tumor sections, were mapped in their entirety using a 250-µm-square measurement grid. Two or more smaller regions of interest within each tissue were additionally mapped using a 25 µm-square step size. A deep learning algorithm, convolutional neural network (CNN), was developed to distinguish histopathologic features with-in individual and across multiple tissue sections. Cancerous breast tissue were discriminated from normal breast tissue with 90 % accuracy, 88.8 % sensitivity and 90.8 % specificity with an excellent Area Under the Receiver Operator Curve (AUROC) of 0.96. Features that contributed significantly to the model were identified and used to generate RGB images of the tissue sections. For each grid point (pixel) on a Raman map, color was assigned to intensities at frequencies of 1002 cm-1 (Phenylalanine), 869 cm-1 (Proline, CC stretching of hydroxyproline-collagen assignment, single bond stretching vibrations for the amino acids proline, valine and polysaccharides) and 1309 cm-1 (CH3/CH2 twisting or bending mode of lipids). The Raman images clearly associate with hematoxylin and eosin stained tissue sections and allow clear visualization of boundaries between normal adipose, connective tissue and tumor. We demonstrated that this simple imaging technique allows high-resolution, straightforward molecular interpretation of Raman images. Raman spectroscopy provides rapid, label-free imaging of microscopic features with high accuracy. This method has application as laboratory tool and can assist with intraoperative tissue assessment during Breast Conserving surgery.

Doxycycline-loaded coaxial nanofiber coating of titanium implants enhances osseointegration and inhibits Staphylococcus aureus infection.

Few studies have been reported that focus on developing implant surface nanofiber (NF) coating to prevent infection and enhance osseointegration by local drug release. In this study, coaxial doxycycline (Doxy)-doped polycaprolactone/polyvinyl alcohol (PCL/PVA) NFs were directly deposited on a titanium (Ti) implant surface during electrospinning. The interaction of loaded Doxy with both PVA and PCL NFs was characterized by Raman spectroscopy. The bonding strength of Doxy-doped NF coating on Ti implants was confirmed by a stand single-pass scratch test. The improved implant osseointegration by PCL/PVA NF coatings in vivo was confirmed by scanning electron microscopy, histomorphometry and micro computed tomography (µCT) at 2, 4 and 8 weeks after implantation. The bone contact surface (%) changes of the NF coating group (80%) is significantly higher than that of the no NF group (<5%, p < 0.05). Finally, we demonstrated that a Doxy-doped NF coating effectively inhibited bacterial infection and enhanced osseointegration in an infected (Staphylococcus aureus) tibia implantation rat model. Doxy released from NF coating inhibited bacterial growth up to 8 weeks in vivo. The maximal push-in force of the Doxy-NF coating (38 N) is much higher than that of the NF coating group (6.5 N) 8 weeks after implantation (p < 0.05), which was further confirmed by quantitative histological analysis and µCT. These findings indicate that coaxial PCL/PVA NF coating doped with Doxy and/or other drugs have great potential in enhancing implant osseointegration and preventing infection.

Shining light on neurosurgery diagnostics using Raman spectroscopy.

Surgical excision of brain tumors provides a means of cytoreduction and diagnosis while minimizing neurologic deficit and improving overall survival. Despite advances in functional and three-dimensional stereotactic navigation and intraoperative magnetic resonance imaging, delineating tissue in real time with physiological confirmation is challenging. Raman spectroscopy is a promising investigative and diagnostic tool for neurosurgery, which provides rapid, non-destructive molecular characterization in vivo or in vitro for biopsy, margin assessment, or laboratory uses. The Raman Effect occurs when light temporarily changes a bond's polarizability, causing change in the vibrational frequency, with a corresponding change in energy/wavelength of the scattered photon. The recorded inelastic scattering results in a "fingerprint" or Raman spectrum of the constituent under investigation. The amount, location, and intensity of peaks in the fingerprint vary based on the amount of vibrational bonds in a molecule and their ensemble interactions with each other. Distinct differences between various pathologic conditions are shown as different intensities of the same peak, or shifting of a peak based on the binding conformation. Raman spectroscopy has potential for integration into clinical practice, particularly in distinguishing normal and diseased tissue as an adjunct to standard pathologic diagnosis. Further, development of fiber-optic Raman probes that fit through the instrument port of a standard endoscope now allows researchers and clinicians to utilize spectroscopic information for evaluation of in vivo tissue. This review highlights the need for such an instrument, summarizes neurosurgical Raman work performed to date, and discusses the future applications of neurosurgical Raman spectroscopy.

Identification of regions of normal grey matter and white matter from pathologic glioblastoma and necrosis in frozen sections using Raman imaging.

In neurosurgical applications, a tool capable of distinguishing grey matter, white matter, and areas of tumor and/or necrosis in near-real time could greatly aid in tumor resection decision making. Raman spectroscopy is a non-destructive spectroscopic technique which provides molecular information about the tissue under examination based on the vibrational properties of the constituent molecules. With careful measurement and data processing, a spatial step and repeat acquisition of Raman spectra can be used to create Raman images. Forty frozen brain tissue sections were imaged in their entirety using a 300-µm-square measurement grid, and two or more regions of interest within each tissue were also imaged using a 25 µm-square step size. Molecular correlates for histologic features of interest were identified within the Raman spectra, and novel imaging algorithms were developed to compare molecular features across multiple tissues. In previous work, the relative concentration of individual biomolecules was imaged. Here, the relative concentrations of 1004, 1300:1344, and 1660 cm(-1), which correspond primarily to protein and lipid content, were simultaneously imaged across all tissues. This provided simple interpretation of boundaries between grey matter, white matter, and diseased tissue, and corresponded with findings from adjacent hematoxylin and eosin-stained sections. This novel, yet simple, multi-channel imaging technique allows clinically-relevant resolution with straightforward molecular interpretation of Raman images not possible by imaging any single peak. This method can be applied to either surgical or laboratory tools for rapid, non-destructive imaging of grey and white matter.

Raman molecular imaging of brain frozen tissue sections.

Raman spectroscopy provides a molecular signature of the region being studied. It is ideal for neurosurgical applications because it is non-destructive, label-free, not impacted by water concentration, and can map an entire region of tissue. The objective of this paper is to demonstrate the meaningful spatial molecular information provided by Raman spectroscopy for identification of regions of normal brain, necrosis, diffusely infiltrating glioma and solid glioblastoma (GBM). Five frozen section tissues (1 normal, 1 necrotic, 1 GBM, and 2 infiltrating glioma) were mapped in their entirety using a 300-µm-square step size. Smaller regions of interest were also mapped using a 25-µm step size. The relative concentrations of relevant biomolecules were mapped across all tissues and compared with adjacent hematoxylin and eosin-stained sections, allowing identification of normal, GBM, and necrotic regions. Raman peaks and peak ratios mapped included 1003, 1313, 1431, 1585, and 1659 cm(-1). Tissue maps identified boundaries of grey and white matter, necrosis, GBM, and infiltrating tumor. Complementary information, including relative concentration of lipids, protein, nucleic acid, and hemoglobin, was presented in a manner which can be easily adapted for in vivo tissue mapping. Raman spectroscopy can successfully provide label-free imaging of tissue characteristics with high accuracy. It can be translated to a surgical or laboratory tool for rapid, non-destructive imaging of tumor margins.

Emerging technology: applications of Raman spectroscopy for prostate cancer.

There is a need in prostate cancer diagnostics and research for a label-free imaging methodology that is nondestructive, rapid, objective, and uninfluenced by water. Raman spectroscopy provides a molecular signature, which can be scaled from micron-level regions of interest in cells to macroscopic areas of tissue. It can be used for applications ranging from in vivo or in vitro diagnostics to basic science laboratory testing. This work describes the fundamentals of Raman spectroscopy and complementary techniques including surface enhanced Raman scattering, resonance Raman spectroscopy, coherent anti-Stokes Raman spectroscopy, confocal Raman spectroscopy, stimulated Raman scattering, and spatially offset Raman spectroscopy. Clinical applications of Raman spectroscopy to prostate cancer will be discussed, including screening, biopsy, margin assessment, and monitoring of treatment efficacy. Laboratory applications including cell identification, culture monitoring, therapeutics development, and live imaging of cellular processes are discussed. Potential future avenues of research are described, with emphasis on multiplexing Raman spectroscopy with other modalities.

Raman spectroscopy to distinguish grey matter, necrosis, and glioblastoma multiforme in frozen tissue sections.

The need exists for a highly accurate, efficient and inexpensive tool to distinguish normal brain tissue from glioblastoma multiforme (GBM) and necrosis boundaries rapidly, in real-time, in the operating room. Raman spectroscopy provides a unique biochemical signature of a tissue type, with the potential to provide intraoperative identification of tumor and necrosis boundaries. We aimed to develop a database of Raman spectra from normal brain, GBM, and necrosis, and a methodology for distinguishing these pathologies. Raman spectroscopy was used to measure 95 regions from 40 frozen tissue sections using 785 nm excitation wavelength. Review of adjacent hematoxylin and eosin sections confirmed histology of each region. Three regions each of normal grey matter, necrosis, and GBM were selected as a training set. Ten regions were selected as a validation set, with a secondary validation set of tissue regions containing freeze artifact. Grey matter contained higher lipid (1061, 1081 cm(-1)) content, whereas necrosis revealed increased protein and nucleic acid content (1003, 1206, 1239, 1255-1266, 1552 cm(-1)). GBM fell between these two extremes. Discriminant function analysis showed 99.6, 97.8, and 77.5% accuracy in distinguishing tissue types in the training, validation, and validation with freeze artifact datasets, respectively. Decreased classification in the freeze artifact group was due to tissue preparation damage. This study shows the potential of Raman spectroscopy to accurately identify normal brain, necrosis, and GBM as a tool to augment pathologic diagnosis. Future work will develop mapped images of diffuse glioma and neoplastic margins toward development of an intraoperative surgical tool.

Conclusions and data analysis: a 6-year study of Raman spectroscopy of solid tumors at a major pediatric institute.

PURPOSE: Create a Raman spectroscopic database with potential to diagnose cancer and investigate two different diagnostic methodologies. Raman spectroscopy measures the energy of photons scattered inelastically by molecules. These molecular signatures form the basis of identifying complex biomolecules and can be used to differentiate normal from neoplastic tissue. METHODS: 1,352 spectra from 55 specimens were collected from fresh or frozen normal brain, kidney and adrenal gland and their malignancies. Spectra were obtained utilizing a Renishaw Raman microscope (RM1000) at 785 nm excitation wavelength with an exposure time of 10 to 20 s/spectrum over three accumulations. Spectra were preprocessed and discriminant function analysis was used to classify spectra based on pathological gold standard. RESULTS: The results of leave 25 % out training/testing validation were as follows: 94.3 % accuracy for training and 91.5 % for testing adrenal, 95.1 % accuracy for training and 88.9 % for testing group of brain, and 100 % accuracy for kidney training/testing groups when tissue origin was assumed. A generalized database not assuming tissue origin provided 88 % training and 85.5 % testing accuracy. CONCLUSION: A database can be made from Raman spectra to classify and grade normal from cancerous tissue. This database has the potential for real time diagnosis of fresh tissue and can potentially be applied to the operating room in vivo.

Identification of pediatric brain neoplasms using Raman spectroscopy.

PURPOSE: Raman spectroscopy can quickly and accurately diagnose tissue in near real-time. This study evaluated the capacity of Raman spectroscopy to diagnose pediatric brain tumors. EXPERIMENTAL DESIGN: Samples of untreated pediatric medulloblastoma (4 samples and 4 patients), glioma (i.e. astrocytoma, oligodendroglioma, ependymoma, ganglioglioma and other gliomas; 27 samples and 19 patients), and normal brain samples (33 samples and 5 patients) were collected fresh from the operating room or from our frozen tumor bank. Samples were divided and tested using routine pathology and Raman spectroscopy. Twelve Raman spectra were collected per sample. Support vector machine analysis was used to classify spectra using the pathology diagnosis as the gold standard. RESULTS: Normal brain (321 spectra), glioma (246 spectra) and medulloblastoma (82 spectra) were identified with 96.9, 96.7 and 93.9% accuracy, respectively, when compared with each other. High-grade ependymomas (41 spectra) were differentiated from low-grade ependymomas (25 spectra) with 100% sensitivity and 96.0% specificity. Normal brain tissue was distinguished from low-grade glioma (118 spectra) with 91.5% sensitivity and 97.8% specificity. For these analyses, the tissue-level classification was determined to be 100% accurate. CONCLUSION: These results suggest Raman spectroscopy can accurately distinguish pediatric brain neoplasms from normal brain tissue, similar tumor types from each other and high-grade from low-grade tumors.

Differentiation of small round blue cell tumors using Raman spectroscopy.

PURPOSE: Small round blue cell tumors (SRBCTs) are aggressive undifferentiated embryonal tumors, including neuroblastoma, rhabdomyosarcoma, Ewing sarcoma, and non-Hodgkin lymphoma. They share similar histologic features. Additional studies such as immunohistochemistry and molecular techniques are required to differentiate them. There is no widely available tool for real-time diagnosis. Raman spectroscopy is an analytical technique with potential for quick and accurate diagnosis of tumors in near real-time. METHODS: Fresh or banked frozen tissue samples from SRBCTs were processed for routine pathology and Raman spectroscopy. Raman results were correlated with the final pathology diagnosis. RESULTS: The data set was composed of 480 spectra from 32 samples, including 179 neuroblastoma, 37 Ewing sarcoma, 164 rhabdomyosarcoma, and 100 non-Hodgkin lymphoma spectra. Discriminant function analysis showed that a combination of 18 peaks could accurately identify 94% of spectra. At the tissue level, all tumors were correctly identified. Only 10 peaks were needed to classify all tissues with 100% accuracy. Spectral-level classification with this model was 87.9%. CONCLUSION: Raman spectroscopy is an accurate technique for quickly and accurately differentiating SRBCTs. It could diagnose these specimens and provide a single, easy to use test for near real-time diagnosis.

Diagnosis of Wilms' tumor using near-infrared Raman spectroscopy.

PURPOSE: Raman spectroscopy has distinguished malignant from normal tissues in several types of cancer. This is the first report of applying Raman spectroscopy to the diagnosis of Wilms' tumor. METHODS: Specimens of normal kidney, Wilms' tumor, xanthogranuloma, nephrogenic rests, and rhabdoid tumor were collected fresh from the operating room. Specimens of Wilms' tumor, normal kidney, and congenital mesoblastic nephroma were retrieved from the cryobank and thawed to room temperature. At least 12 Raman spectra were collected from each tissue sample. Histologic slides of each specimen were reviewed by pediatric pathologists. A computer algorithm based on discriminant function analysis (DFA) classified the Raman spectra of Wilms' tumor and the normal sample. RESULTS: Four hundred sixty-seven spectra were collected from 41 specimens. Using DFA, Raman spectroscopy differentiated Wilms' tumor from normal with 100% sensitivity and specificity and treated from untreated Wilms' tumor with 100% sensitivity and specificity. Using a DFA model built from cryopreserved specimens but applied to fresh Wilms' and normal samples, the sensitivity and specificity were 93.3% and 90.9%, respectively. CONCLUSION: Raman spectroscopy is an accurate technique for differentiating Wilms' tumor from normal kidney and treated from untreated Wilms' tumor. It has potential to diagnose in minutes what currently takes several hours to days.

Raman spectroscopy detects and distinguishes neuroblastoma and related tissues in fresh and (banked) frozen specimens.

BACKGROUND: Raman spectroscopy has been shown to accurately distinguish different neural crest-derived pediatric tumors. This study tests the ability of Raman spectroscopy to accurately identify cryopreserved tissue specimens using a classification algorithm designed from fresh tumor data and vice versa. METHODS: Fresh specimens of neuroblastoma and other pediatric neural crest tumors were analyzed with Raman spectroscopy. After analysis, the specimens were stored at -80 degrees C. At a later date, the specimens were thawed and reanalyzed by Raman spectroscopy. A computer algorithm was used to classify the spectra from the frozen tissue against a computer model built on the fresh tissue data. This classification process was then reversed, testing fresh spectra against a model built from frozen data. RESULTS: We collected 1114 spectra (862 fresh and 252 frozen) from 62 tissue samples, including 8 normal adrenal glands, 29 neuroblastomas, 14 ganglioneuromas, 8 nerve sheath tumors, and 3 pheochromocytomas. At the tissue level, frozen neuroblastoma, ganglioneuroma, nerve sheath tumor, and pheochromocytoma were distinguished from normal adrenal tissue with 100% sensitivity and specificity. Fresh tissue had the same results except for the misclassification of one specimen of nerve sheath tumor. CONCLUSIONS: The representative spectra show a high correlation between fresh and frozen tissue, and a clear difference between pathologic conditions. Spectra from frozen tissue can be accurately classified against spectra from fresh tissue and vice versa. This modality makes it possible to determine in a few minutes a result that often takes 12 to 36 hours for tissue processing and consideration by a trained pathologist to achieve.

Evaluation of pancreatic cancer with Raman spectroscopy in a mouse model.

OBJECTIVES: Detection of neoplastic changes using optical spectroscopy has been an active area of research in recent times. Raman spectroscopy is a vibrational spectroscopic technique that can be used to diagnose various tumors with high sensitivity and specificity. We evaluated the ability of Raman spectroscopy to differentiate normal pancreatic tissue from malignant tumors in a mouse model. METHODS: We collected 920 spectra, 475 from 31 normal pancreatic tissue and 445 from 29 tumor nodules using a 785-nm near-infrared laser excitation. Discriminant function analysis was used for classification of normal and tumor samples. RESULTS: Using principal component analysis, we were able to highlight subtle chemical differences in normal and malignant tissue. Using histopathology as the gold standard, Raman analysis gave sensitivities between 91% and 96% and specificities between 88% and 96%. CONCLUSIONS: Raman spectroscopy along with discriminant function analysis is a useful method to detect cancerous changes in the pancreas. Pancreatic tumors were characterized by increased collagen content and decreased DNA, RNA, and lipids components compared with normal pancreatic tissue.

Raman spectroscopy can differentiate malignant tumors from normal breast tissue and detect early neoplastic changes in a mouse model.

Raman spectroscopy shows potential in differentiating tumors from normal tissue. We used Raman spectroscopy with near-infrared light excitation to study normal breast tissue and tumors from 11 mice injected with a cancer cell line. Spectra were collected from 17 tumors, 18 samples of adjacent breast tissue and lymph nodes, and 17 tissue samples from the contralateral breast and its adjacent lymph nodes. Discriminant function analysis was used for classification with principal component analysis scores as input data. Tissues were examined by light microscopy following formalin fixation and hematoxylin and eosin staining. Discriminant function analysis and histology agreed on the diagnosis of all contralateral normal, tumor, and mastitis samples, except one tumor which was found to be more similar to normal tissue. Normal tissue adjacent to each tumor was examined as a separate data group called tumor bed. Scattered morphologically suspicious atypical cells not definite for tumor were present in the tumor bed samples. Classification of tumor bed tissue showed that some tumor bed tissues are diagnostically different from normal, tumor, and mastitis tissue. This may reflect malignant molecular alterations prior to morphologic changes, as expected in preneoplastic processes. Raman spectroscopy not only distinguishes tumor from normal breast tissue, but also detects early neoplastic changes prior to definite morphologic alteration.