Assessment of multiple signaling pathway activity scores in airway epithelial cells to predict airway epithelial abnormalities and risk of airway cancer
A computer-based method analyzing signaling pathways in airway epithelial cells addresses the limitations of current lung cancer detection by accurately identifying pre-malignant changes and cancer risk, reducing invasive procedures and radiation exposure.
Patent Information
- Application Number
- JP2021567791
- Authority / Receiving Office
- JP · JP
- Patent Type
- Patents
- Current Assignee / Owner
- Priority Date
- 2019-05-13
- Filing Date
- 2020-05-08
- Publication Date
- 2024-11-15
- Estimated Expiration
- 2040-05-08
AI Technical Summary
Current lung cancer detection methods, such as LDCT screening, have high false-positive rates and expose patients to radiation, while invasive biopsies are risky and often inconclusive, necessitating a more sensitive and specific method to identify early airway epithelial abnormalities and cancer risk.
A computer-implemented method analyzing the activity of cell signaling pathways, specifically TGF-β, PI3K-FOXO, and Notch pathways, in airway epithelial cells to determine an airway abnormality factor or risk score indicative of pre-malignant changes and cancer risk, using a calibrated mathematical model to interpret pathway activities.
This approach allows for early identification of individuals at high risk of lung cancer, reducing unnecessary invasive procedures and radiation exposure by providing accurate differentiation between benign and malignant nodules, and guiding targeted therapies.
Smart Images

Figure 0007583999000009 
Figure 0007583999000010 
Figure 0007583999000011
Abstract
Description
[Technical field]
[0001] The subject matter described herein primarily relates to bioinformatics, genomic processing, proteomic processing, and related technologies. More particularly, the present invention relates to a computer-implemented method for determining whether a subject has abnormal airway epithelium, and a computer-implemented method for determining a risk score indicative of the risk of a subject with abnormal airway epithelium developing airway cancer. The present invention further relates to an apparatus, a non-transitory storage medium, and a computer program for determining an airway abnormality factor indicative of whether a subject has abnormal airway epithelium or a risk score indicative of the risk of a subject with abnormal airway epithelium developing airway cancer. The present invention further relates to a kit for determining an airway abnormality factor indicative of whether a subject has abnormal airway epithelium or a risk score indicative of the risk of a subject with abnormal airway epithelium developing airway cancer, as well as a kit for use in the corresponding diagnostic or prognostic method. The airway abnormality factor and the risk score are determined based on a combination of signaling pathway activities. [Background technology]
[0002] Lung cancer is a fatal disease with poor prognosis and is the leading cause of cancer deaths worldwide. Many patients are diagnosed at advanced stages, with a 5-year survival rate of only 18.6%. Early detection and timely diagnosis, staging, and treatment are therefore important in the treatment of lung cancer. Screening programs are sometimes oriented towards early detection of lung cancer in high-risk groups, aiming to improve clinical outcomes by administering treatment at an early stage. Low-dose CT (LDCT) screening of high-risk populations has been performed, where eligibility for screening is primarily based on smoking history. However, one of the drawbacks of LDCT screening is the relatively high false-positive rate (NLST Research Team. N Engl J Med;368:1980-1991 (2013)), as well as exposure to cumulative radiation doses when screening procedures are repeated. Positive CT findings require further follow-up with potentially unnecessary biopsy procedures to obtain pathological confirmation, with the attendant health risks to the patient at hand. Depending on the location of the tumor, obtaining tumor samples for diagnosis / subtyping can be difficult. Thus, there is a need for a test for early detection of lung cancer with high sensitivity / specificity and without radiation exposure. To reduce the number of unnecessary CT scans and the resulting false positives, there is also a need for a minimally invasive test to identify high-risk populations for subsequent next step image analysis. Also, when a nodule is found, for example, on a CT scan, a decision needs to be made as to whether a biopsy needs to be performed to allow a pathological diagnosis, which is an invasive and risky procedure. Thus, there is a need for a method to help make this decision by providing complementary evidence regarding the possible cause of the identified nodule, i.e. benign or malignant.
[0003] The most common type of lung cancer is non-small cell lung cancer (NSCLC), of which the most common subtypes are adenocarcinoma (40%) and squamous cell carcinoma (30%). Smoking is responsible for approximately 85% of all lung cancers. Roughly 5-15% of smokers develop lung cancer. Patients with COPD associated with (heavy) smoking are at increased risk of developing lung cancer (Oncotarget. 2017 Sep 29;8(44):78044-78056; Lung Cancer. 2015 Nov;90(2):121-127; Respiration. 2011;81(4):265-84).
[0004] Tobacco exposure is an important factor in inducing premalignant changes. The majority of lung cancers are associated with smoking, but 15% occur in never smokers. Heavy smoking is associated with an increased risk of lung cancer, specifically squamous cell and small cell types, as well as other histopathological types.
[0005] Adenocarcinoma is the predominant subtype in never smokers, and squamous cell carcinoma is rare in never smokers. Adenocarcinomas often have a peripheral or endobronchial origin, while squamous cell carcinomas usually arise in the trachea or proximal airways.
[0006] Tobacco exposure can induce stress in airway epithelial cells, leading to smoking-induced injury that requires epithelial regeneration and repair. Smoking interferes with signaling pathways involved in these processes, which can promote tumorigenesis.
[0007] Common premalignant lesions, for example in colon adenocarcinomas, but also in airway epithelia, are characterized by increased proliferation reflected by the aberrant activity of certain signal transduction pathways, such as the PI3K pathway (Clin Cancer Res. 2018 Jul 1; 24(13): 2984-2992, Sci Transl Med. 2010 Apr 7; 2(26): 26ra25). Aberrant proliferation can be enabled by the activity of one or more growth factor pathways, such as the PI3K pathway or the MAPK-AP1 or JAK-STAT pathways, with the loss of pathway activity that inhibits / controls cell proliferation, such as the TGF-β or Notch pathways in normal cells. During progression to cancer, other tumorigenic signal transduction pathways are recruited or more regulatory pathways are lost due to an increasing number of genomic mutations and / or chromosomal abnormalities. That is, premalignant lesions may share some pathway activity characteristics, but not necessarily all, with the final cancer. Perhaps for this reason, little is known about the role of signal transduction pathways in the development of (lung) cancer and the interference of smoking with pathway activity.
[0008] Obtaining epithelial cell samples from large or small airways for analysis: Analysis of samples from the epithelial lining of the airways can provide information about pre-malignant changes or provide further diagnostic information if, for example, a suspicious nodule is seen on a CT scan.
[0009] Lung cancer develops in the epithelium of large airways (bronchi that branch off the trachea) or small airways. Using techniques such as airway scraping or bronchoalveolar lavage, epithelial cells can be obtained for molecular analysis from the upper and lower airways in a relatively non-invasive manner. This provides a potential means of identifying individuals with abnormal tumorigenic signaling pathway activity in the airway epithelium that is indicative of early proliferative changes in the airway epithelium and at increased risk for developing lung cancer, i.e. smokers. COPD patients are generally chronic heavy smokers and constitute a high-risk group for the development of lung cancer.
[0010] Aberrant signal transduction pathway activity, if identified, could in principle be regulated by targeted drugs targeting specific signaling pathways, e.g., blocked by PI3K pathway inhibitors (W Verhaegh et al., Cancer research, 2014; 74(11):2936-45; A van de Stolpe et al., Scientific Reports, 2019, 9(1603); van Ooijen, Am J Pathol, 2018, 188(9):1956-1972). This opens the option that pre-malignant airway epithelial lesions are clinically treatable and abnormalities can be reversed or eliminated by local treatment of the airway epithelium. Summary of the Invention [Problem to be solved by the invention]
[0011] Thus, there is a high need for a method to detect early changes in airway epithelium indicative of abnormal epithelium and characterize abnormalities related to abnormal signaling pathway activity. The method is expected to be used for the identification of individuals at high risk of lung cancer, provide further evidence for the malignant and benign nature of identified lung nodules, and indicate which therapies are useful for treating identified lesions or associated lung cancer. [Means for solving the problem]
[0012] According to a first aspect of the present invention, the above problem is solved by a computer-implemented method performed by a digital processing device for determining whether a subject has abnormal airway epithelium, the determining step includes a step of determining an airway abnormality factor indicative of whether the subject has abnormal airway epithelium based on a combination of activity of cell signaling pathways in an epithelial cell sample from the subject's airway, the cell signaling pathways including two or more cell signaling pathways selected from the group consisting of TGF-β pathway, PI3K-FOXO pathway, and Notch pathway, and the step of determining the signaling pathway abnormality factor is further based on a reference activity of each cell signaling pathway, the reference activity reflecting the activity of each cell signaling pathway found in the airway epithelium of a healthy subject. Additional cell signaling pathways may be optionally included in the method of the present invention. In one embodiment, the method further includes a step of providing an airway abnormality factor for any one of the various uses disclosed herein, such as determining a risk score indicative of the risk of a subject with abnormal airway epithelium to develop airway cancer, defining an interval for repeating the method thereafter, recommending supplementary diagnostics to be performed, recommending treatment, etc.
[0013] According to a second aspect of the present invention, the problem is solved by a computer-implemented method performed by a digital processing device for determining a risk score indicative of the risk of a subject having abnormal airway epithelium to develop airway cancer, the determining step comprising determining the risk score based on a combination of cell signaling pathway activities in an epithelial cell sample from the subject's airway, the analysis of the cell signaling pathways comprising two or more cell signaling pathways selected from the group consisting of the TGF-β pathway, the PI3K-FOXO pathway, and the Notch pathway, and the determining the risk score is further based on a combination of reference activities of the cell signaling pathways, the risk score being defined such that the indicated risk increases with one or more of a decrease in the activity of the TGF-β pathway and an increase in the activity of the PI3K pathway and / or a decrease in the activity of the Notch pathway relative to the reference activities of the cell signaling pathways. In one embodiment, the method further comprises providing a risk score for any one of the various applications disclosed herein, such as defining an interval at which the method should then be repeated, recommending an auxiliary diagnostic method to be performed, recommending a treatment, etc.
[0014] The present invention is based on the inventors' innovation that the analysis of signal transduction pathway activity can be used to characterize even the pre-malignant risk of airway epithelium, which plays a role in the progression to malignant airway cancer.The inventors have found that the presence or absence of malignant and pre-malignant risk can be evaluated by measuring the activity of certain signal transduction pathways, and have evaluated the measured activities in combination.The inventors are the first to provide insight into how related pathway activities are interrelated in determining whether airway epithelium is abnormal and / or whether abnormal airway epithelium characterizes subjects at risk of developing airway cancer.
[0015] The present invention was achieved by focusing on the activity of signaling pathways (especially TGF-β, PI3K-FOXO, Notch pathways) in airway epithelial cells of subjects belonging to various groups, including healthy non-smokers, light smokers, and heavy smokers. A computer model was then developed for the interpretation of the measured pathway activity to determine abnormal deviations from normal pathway activity that can be exhibited by signaling pathway abnormality factors, and to provide airway abnormality factors and / or risk scores for subjects with unknown characteristics of airway epithelium. In this model, the combined activity of signaling pathways is used as an indicator (biomarker) to characterize the possibility that a subject has abnormal airway epithelium and / or is at risk for developing airway cancer.
[0016] As an advantage, the present invention facilitates the identification of subjects at risk of developing cancer at an early stage of the disease, thus enabling better treatment options than existing methods. It is expected that the use of the present method will reduce the number of unnecessary CT scans, thereby reducing the number of unnecessary diagnostic procedures and exposure to harmful radiation. The development of lung cancer is a multi-step process in which normal lung epithelial cells accumulate genetic abnormalities and are transformed into a malignant phenotype. The method of the present invention has great potential to aid in the detection of early (pre-malignant) changes that are ultimately at risk of becoming malignant at a later stage, and in this context can be applied in various clinical scenarios. Before airway cancer is detectable, the method of the present invention can be used for risk assessment and identification of populations that would benefit from screening or primary chemoprevention, or local treatments that block progression or reverse the lesion to normal. Furthermore, the method of the present invention can be used to detect lung cancer at pre-malignant or early stages before it progresses to invasive lung cancer as mentioned. The method of the present invention can be an aid in the selection of individuals for close CT monitoring and secondary chemoprevention. Furthermore, the method of the present invention has the value of differentiating and predicting benign and malignant nodules found in CT images. Finally, some biopsy procedures have limited sensitivity and therefore do not always result in a confirmed diagnosis. For example, bronchoscopy is often non-diagnostic and may require further invasive testing. By combining such known diagnostic tests with the method of the present invention, it is expected that their diagnostic performance will be significantly improved. For example, when a non-diagnostic nodule is found in a CT scan, the method will help classify the nodule as malignant or benign.
[0017] The term "subject" as used herein means any living organism. In some embodiments, the subject is an animal, preferably a mammal. In certain embodiments, the subject is a human, e.g., a medical subject. The applicability of the method of the present invention is not limited to a particular group of subjects, but it will be clear that subjects belonging to high-risk groups, such as smokers or COPD, will benefit most from the present invention. Thus, the subject to be diagnosed is preferably a smoker, in particular a tobacco smoker, or a subject who is or has been exposed to smoke on a daily basis, e.g., a subject who lives in the same household as a smoker. Other risk factors that preferably define the subject to be diagnosed include exposure to radon, asbestos, arsenic, diesel exhaust, silica, and chromium. The method of the present invention is advantageously applied repeatedly, especially at regular intervals, so that premalignant changes in the airway epithelium can be detected as early as possible.
[0018] The epithelial cell sample to be used according to the invention may be an extracted sample, i.e. a sample extracted from a subject. Examples of samples include, but are not limited to, epithelial cells, tissues, and / or body fluids containing epithelial cells from the respiratory tract of a subject. The epithelial cell sample may also be a sample containing progenitor cells and / or stem cells of epithelial cells, e.g. basal cells (BC), which constitute the stem / progenitor cells required for the regeneration of damaged epithelial cells. The respective sample may be obtained from the upper or lower respiratory tract of a subject, e.g. by bronchoalveolar lavage, scraping, biopsy, etc. The term "sample" as used herein also encompasses cases where, e.g., cells, tissues, and / or body fluids are removed from a subject and placed, e.g., on a microscope slide or mounting platform, and a portion of this sample is extracted, e.g., by laser capture microdissection (LCM), or punching, or exfoliation of cells of interest from the slide, or by fluorescence activated cell sorting techniques, to carry out the claimed method. Additionally, the term "sample" as used herein encompasses instances where, for example, cells, tissues, and / or bodily fluids are removed from a subject and placed, for example, on a microscope slide, and the claimed methods are performed on the slide.
[0019] The term "upper respiratory tract" as used herein includes the nasal cavity, pharynx (including the nasopharynx, oropharynx, and laryngopharynx), and larynx. The term "lower respiratory tract" as used herein includes the trachea and lungs (including the primary bronchi and bronchioles).
[0020] The terms "pathway," "signal transduction pathway," "signal transduction pathway," and "cell signaling pathway" are used interchangeably herein.
[0021] "Signal pathway activity" refers to the activity of a transcription factor (TF) element associated with a signal transduction pathway in a sample that controls the transcription of a target gene in driving expression of the target gene, i.e., the rate at which the target gene is transcribed, e.g., in terms of high activity (i.e., fast rate) or low activity (i.e., slow rate), or other dimensions such as the level, value, etc. associated with such activity (e.g., rate). Thus, for purposes of the present invention, the term "activity" as used herein is also intended to refer to activity levels obtained as intermediate results during the "pathway analysis" described herein.
[0022] The term "transcription factor element" (TF element), as used herein, preferably refers to an intermediate or precursor protein or protein complex of an active transcription factor, or an active transcription factor protein or protein complex that controls the expression of a specified target gene. For example, the protein complex comprises at least the intracellular domain of one of the respective signal transduction pathway proteins and one or more cofactors, thereby controlling the transcription of the target gene. Preferably, the term refers to a protein or protein complex transcription factor that is induced by cleavage of one of the respective signal transduction pathway proteins that results in the intracellular domain.
[0023] The term "target gene" as used herein means a gene whose transcription is directly or indirectly controlled by a respective transcription factor element. A "target gene" may be a "direct target gene" and / or an "indirect target gene" (as described herein).
[0024] The terms "abnormal" and "abnormality" as used herein refer to characteristics attributed to a subject, cell, or cell sample that are considered rare, dysfunctional, poorly functioning, or malignant, and in particular characterize "pre-malignant" changes in a subject, cell, or cell sample. Diagnosis of abnormal (pre-malignant) characteristics indicates that the subject is at risk for malignant airway cancer. For the purposes of the present invention, a subject, cell, or cell sample is considered abnormal if it is determined to deviate from a respective reference pathway activity based on the respective pathway activity disclosed herein. The reference pathway activity is particularly the pathway activity found in a healthy (i.e., normal) subject, a cell derived from a healthy subject, or a sample comprising cells derived from a healthy subject. If not already known, the reference pathway activity can be experimentally determined by the pathway analysis disclosed herein using epithelial cell samples from one or more healthy subjects. Preferably, samples from multiple healthy subjects are evaluated taking into account natural pathway activity variations.
[0025] According to this, it is an embodiment of the present invention that the determination of the airway abnormality factor and / or risk score is further based on the respective combination of reference activity of the signaling pathway.Similarly, the determination of the signaling pathway abnormality factor is further based on the reference activity of each signaling pathway.The reference activity reflects the activity of each signaling pathway found in the airway epithelium of healthy subjects.
[0026] By comparing each of the reference pathway activities with each of the pathway activities in the subject to be diagnosed, a signaling pathway abnormality factor can be determined for each of the respective pathways. The signaling pathway abnormality factor indicates whether the activity of each of the pathways deviates (abnormally) from the reference activity of each of the pathways. The signaling pathway abnormality factor is then converted into an airway abnormality factor. The airway abnormality factor can also be calculated directly from the combination of pathway activities. The airway abnormality factor can also be considered as a multiple pathway score, i.e., MPS, and indicates the possibility that a subject has abnormal airway epithelium. Thus, "airway abnormality factor", also referred to herein as "airway pathway abnormality score" (APAS) or specifically "small airway pathway abnormality score" (SAPAS), or "large airway pathway abnormality score" (LAPAS), refers to a dimension, e.g., a level or value, that relates a combination of pathway activities to the possibility that a subject has abnormal (e.g., small or large) airway epithelium.
[0027] The term "airway cancer" as used herein refers to malignant tumors of the airway, and in this context, particularly refers to types of cancer that are at increased risk when smoking. The method can be broadly applied to these types of cancer, especially when they originate from epithelial cells. The present invention focuses on cancers that originate from epithelial cells of the airway. Non-limiting examples of such cancers include tracheal cancer, bronchial cancer, cancer of the upper airway (including nose, mouth, larynx, etc.), lung cancer, subtypes of lung cancer including non-small cell lung cancer (NSCLC) and small cell lung cancer (SCLC), histological subtypes such as squamous cell lung cancer, adenosquamous cell carcinoma, large cell carcinoma, sarcomatoid carcinoma, and lung adenocarcinoma. The risk score can be determined by comparing each of the reference pathway activities with each of the pathway activities in the epithelial cell sample of the subject to be diagnosed. The risk score can also be considered as a multiple pathway score, or MPS, and indicates the likelihood that the subject will develop airway cancer. Thus, a "risk score" refers to a dimension, eg, a level or value, that relates a combination of pathway activity to the likelihood that a subject will develop airway cancer.
[0028] The airway abnormality factors and / or risk factors are based on a "combination of cell signaling pathway activity". This means that the airway abnormality factors and / or risk factors are influenced by the activity of two or more cell signaling pathways. The activity of two or more cell signaling pathways can be predicted and / or combined by the mathematical models described herein. In a preferred embodiment, the airway abnormality factors and / or risk factors are based on a combination of signaling pathway activities including the activity of three or more cell signaling pathways. Such a combination of signaling pathway activities includes the activity of three or four, or five or more, for example, five, six, seven, eight, or more different signaling pathways.
[0029] In general, many different formulas can be devised to determine an airway abnormality factor and / or risk score based on the combined activity of two or more cell signaling pathways in a subject, i.e. MPS=F(Pi)+X, where i=1...N, Here, MPS represents the airway abnormality factor and / or risk score (the term "MPS" is used herein as an abbreviation for "multiple pathway score" to indicate that the risk score is influenced by the activity of more than one cell signaling pathway), Pi represents the activity of the cell signaling pathway, N represents the total number of cell signaling pathways used to calculate the airway abnormality factor and / or risk score, and X is a substitute term for possible additional factors and / or parameters that may be included in the formula. Such a formula may more specifically be a polynomial of a certain degree in a given variable or linear combination of variables. Although the weighting coefficients and multipliers in such a polynomial can be set based on the knowledge of a skilled person, typically a training data set containing known evidence, such as survival data, is used to estimate the weighting coefficients and multipliers of the above formula. The activities can be combined using the above formula, and the MPS is subsequently generated. The weighting coefficients and multipliers of the scoring function are then optimized so that a high MPS correlates with a high probability that a patient has abnormal airway epithelium and / or develops airway cancer, or vice versa. Optimization of the correlation of the scoring function with known data can be performed using a number of analytical techniques, such as the Cox proportional hazards test (preferably used herein), the log-rank test, the Kaplan-Meier estimator in conjunction with standard optimization techniques such as gradient descent or manual fitting, and others.
[0030] According to a preferred embodiment of the present invention, the airway abnormality factor and the risk score are each determined based on the evaluation of a calibrated mathematical model that correlates the activity of the signaling pathways in an epithelial cell sample with the airway abnormality factor and the risk score, respectively. The model is programmed to interpret the combination of pathway activities to determine the airway abnormality factor and / or the risk score of the subject to be diagnosed. In particular, the determination of the airway abnormality factor and / or the risk score includes the steps of (i) receiving the activity of each signaling pathway in an epithelial cell sample of the subject to be diagnosed, and (ii) determining the airway abnormality factor and / or the risk score of the subject, wherein the determining step is based on the evaluation of a calibrated mathematical model that correlates the activity of each signaling pathway with the airway abnormality factor and / or the risk score, respectively.
[0031] The calibrated mathematical pathway model is preferably a centroid model or a linear model, or a Bayesian network model based on conditional probability. For example, the calibrated mathematical pathway model may be a probabilistic model, preferably a Bayesian network model based on conditional probability that associates airway abnormality factors and / or risk scores with signaling pathway activity, or the calibrated mathematical pathway model may be based on one or more linear combinations of signaling pathway activity.
[0032] According to the mathematical model, the activity of the signaling pathway is interpreted to provide an airway abnormality factor which is further converted into a risk score or directly interpreted to provide a risk score. The airway abnormality factor and risk score predict or provide the likelihood that a subject has abnormal airway epithelium and / or that the subject is at risk for developing airway cancer.
[0033] Thus, determining the risk score includes determining an airway abnormality factor based on a combination of the activity of the cell signaling pathways in the epithelial cell sample, and converting the airway abnormality factor into a risk score. Determining the airway abnormality factor includes determining a signaling pathway abnormality factor for each of the respective cell signaling pathways based on the activity of the respective cell signaling pathways in the epithelial cell sample, and determining the airway abnormality factor based on a combination of the determined signaling pathway abnormality factors.
[0034] According to a preferred embodiment of the present invention, the activity of each signaling pathway is determined or determinable by pathway analysis as described herein.
[0035] Pathway analysis allows the quantitative measurement of signal transduction pathway activity in epithelial cells based on inference of signal transduction pathway activity from measurement of mRNA levels of well-validated direct target genes of transcription factors associated with the respective signaling pathway (see e.g. W Verhaegh et al. 2014, supra; W Verhaegh, A van de Stolpe, Oncotarget, 2014, 5(14):5196).
[0036] Preferably, the determination of signaling pathway activity, the combination of multiple pathway activities, and applications thereof are described, for example, in the following documents, each of which is incorporated herein in its entirety for the purpose of determining the activity of a respective signaling pathway: International Patent Application Publication No. WO2013011479 (entitled "ASSESSMENT OF CELLULAR SIGNALING PATHWAY ACTIVITY USING PROBABILISTIC MODELING OF TARGET GENE EXPRESSION"), International Patent Application Publication No. WO2014102668 (entitled "ASSESSMENT OF CELLULAR SIGNALING PATHWAY ACTIVITY USING LINEAR COMBINATION(S) OF TARGET GENE EXPRESSIONS"), International Patent Application Publication No. WO2015101635 (entitled "ASSESSMENT OF THE PI3K CELLULAR SIGNALING PATHWAY ACTIVITY USING MATHEMATICAL MODELLING OF TARGET GENE EXPRESSIONS"), International Patent Application Publication No. WO2013011479 (entitled "ASSESSMENT OF CELLULAR SIGNALING PATHWAY ACTIVITY USING PROBABILISTIC MODEL ... WO2016062891 (title of invention: "ASSESSMENT OF TGF-β CELLULAR SIGNALING PATHWAY ACTIVITY USING MATHEMATICAL MODELLING OF TARGET GENE EXPRESSION"), WO2017029215 (title of invention: "ASSESSMENT OF NFKB CELLULAR SIGNALING PATHWAY ACTIVITY USING MATHEMATICAL MODELLING OF TARGET GENE EXPRESSION"), WO2014174003 (title of invention: "MEDICAL PROGNOSIS AND PREDICTION OF TREATMENT RESPONSE USING MULTIPLE CELLULAR SIGNALLING PATHWAY ACTIVITIES"), WO2016062892 (title of invention: "MEDICAL PROGNOSIS AND PREDICTION OF TREATMENT RESPONSE USING MULTIPLE CELLULAR SIGNALINGWO2016062893 (title of invention: MEDICAL PROGNOSIS AND PREDICTION OF TREATMENT RESPONSE USING MULTIPLE CELLULAR SIGNALING PATHWAY ACTIVITIES) and WO2018096076 (title of invention: Method to distinguish tumor suppressive FOXO activity from oxidative stress) and patent application EP16200697.7 (filed on November 25, 2016, title of invention: Method to distinguish tumor suppressive FOXO activity from oxidative stress) and EP17194288.1 (filed on October 2, 2017, title of invention: Assessment of Notch cellular signaling pathway activity using mathematical modelling of target gene Patents filed on October 2, 2017, entitled "Assessment of JAK-STAT1 / 2 cellular signaling pathway activity using mathematical modelling of target gene expression", EP17194291.5 (filed on October 2, 2017, entitled "Assessment of JAK-STAT3 cellular signaling pathway activity using mathematical modelling of target gene expression"), and EP17209053.2 (filed on December 20, 2017, entitled "Assessment of MAPK-AP1 cellular signaling pathway activity using mathematical modelling of target gene expression").PCT / EP2018 / 076232 (filed on September 27, 2018, titled "Assessment of JAK-STAT3 cellular signaling pathway activity using mathematical modelling of target gene expression"), PCT / EP2018 / 076334 (filed on September 27, 2018, titled "Assessment of JAK-STAT1 / 2 cellular signaling pathway activity using mathematical modelling of target gene expression"), PCT / EP2018 / 076488 (filed on September 28, 2018, titled "Assessment of Notch cellular signaling pathway activity using mathematical modelling of target gene expression"), PCT / EP2018 / 076513 (filed on September 28, 2018, titled "Assessment of MAPK-AP-1 cellular signaling pathway activity using mathematical modelling of target gene expression") The present invention is carried out as described in PCT / EP2018 / 076614 (filed October 1, 2018, entitled "Determining functional status of immune cell types and immune response"), and PCT / EP2018 / 076614 (filed October 1, 2018, entitled "Determining functional status of immune cell types and immune response").
[0037] Models have been biologically validated for the ER, AR, PI3K-FOXO, HH, Notch, TGF-β, Wnt, NFkB, JAK-STAT1 / 2, JAK-STAT3, and MAPK-AP1 pathways in several cell types. It is noted that the mathematical models employed in the as yet unpublished patent applications, and the calibration and use of these models in these applications, generally correspond to the models, calibration, and use disclosed in the previously published patent applications.
[0038] A unique set of cell signaling pathway target genes whose expression levels are preferably analyzed has been identified. Three or more, for example 3, 4, 5, 6, 7, 8, 9, 10, 11, 12 or more target genes from each evaluated cell signaling pathway can be analyzed to determine pathway activity for use in the mathematical model.
[0039] Common to the pathway analysis methods for determining the activity of various signaling pathways disclosed herein is a concept that is preferably applied for the purposes of the present specification, in which the activity of a signaling pathway in a cell, such as an epithelial cell, present in a sample can be determined by receiving the expression level of one or more, preferably three or more, target genes of the signaling pathway, determining the activity level of a transcription factor (TF) element associated with the signaling pathway in the sample that controls the transcription of the three or more target genes, based on evaluation of a calibrated mathematical pathway model that relates the expression level of one or more, preferably three or more, target genes to the activity level of the signaling pathway, and optionally inferring the activity of the signaling pathway in the epithelial cell based on the determined activity level of the TF element associated with the signaling pathway.As described herein, the activity level can be directly used as an input for determining an airway abnormality factor and / or a risk score, which is also contemplated by the present invention.
[0040] The term "activity level" of a TF element, as used herein, refers to the level of activity of a TF element with respect to the transcription of its target gene.
[0041] The calibrated mathematical pathway model is a probabilistic model, preferably a Bayesian network model based on conditional probabilities relating the activity levels of TF elements involved in the signaling pathway and the expression levels of three or more target genes, or the calibrated mathematical pathway model is based on one or more linear combinations of the expression levels of three or more target genes. For the purposes of the present invention, the calibrated mathematical pathway model is preferably a centroid model or a linear model, or a Bayesian network model based on conditional probabilities.
[0042] In particular, the determination of the expression level in the subject and optionally the prediction of the activity of the signaling pathway is performed, for example, inter alia, by (i) evaluating a portion of a calibrated probabilistic pathway model, preferably a Bayesian network, representing the cell signaling pathway for a set of inputs including the expression levels of three or more target genes of the cell signaling pathway measured in a sample of the subject, (ii) predicting the activity level in the subject of a transcription factor (TF) element associated with the signaling pathway that controls the transcription of three or more target genes of the cell signaling pathway based on a conditional probability associated with the activity level of the TF element associated with the signaling pathway and the expression levels of three or more target genes of the cell signaling pathway measured in a sample of the subject, and optionally (iii) predicting the activity of the cell signaling pathway based on the predicted activity level of the TF element associated with the signaling pathway in the sample of the subject. This is described in detail in published international patent application WO2013 / 011479A2 ("Assessment of cellular signaling pathway activity using probabilistic modeling of target gene expression"), the contents of which are incorporated herein in their entirety.
[0043] In an exemplary option, determining the expression level and optionally predicting the activity of the cell signaling pathway in the subject is performed by, inter alia, (i) determining the activity level of a transcription factor (TF) element associated with a signaling pathway in the subject's sample that controls the transcription of three or more target genes of the cell signaling pathway, where the determining step is based on evaluation of a calibrated mathematical pathway model that relates the expression levels of the three or more target genes of the cell signaling pathway to the activity levels of the TF elements associated with the signaling pathway, where the mathematical pathway model is based on one or more linear combinations of the expression levels of the three or more target genes, and optionally (ii) predicting the activity of the cell signaling pathway in the subject based on the determined activity levels of the TF elements associated with the signaling pathway in the subject's sample. This is described in detail in published international patent application WO2014 / 102668A2 ("Assessment of cellular signaling pathway activity using linear combination(s) of target gene expressions").
[0044] Further details regarding predicting cell signaling pathway activity using mathematical modeling of target gene expression can be found in W Verhaegh et al., 2014, supra.
[0045] In one embodiment, the measurement of the signaling pathway is performed using qPCR, multiplex qPCR, multiplex qPCR, ddPCR, RNAseq, RNA expression arrays, or mass spectrometry. For example, gene expression microarray data, such as Affymetrix microarrays, or RNA sequencing methods such as Illumina sequencers can be used.
[0046] The present invention focuses on the TGF-β pathway, the PI3K-FOXO pathway, the Notch pathway, and / or the HH pathway. According to a preferred embodiment of the present invention, the signaling pathway comprises one or more of the TGF-β pathway and the PI3K-FOXO pathway and the Notch pathway, preferably the TGF-β pathway and at least the PI3K-FOXO pathway.
[0047] According to a preferred embodiment of the present invention, the risk score is defined such that the indicated risk increases with one or more of the following: the activity of TGF-β pathway is decreased, and the activity of PI3K pathway is increased, and the activity of Notch pathway is decreased.Similarly, the airway abnormality factor is preferably defined such that the indicated factor reflects an increase in deviation from normality with one or more of the activity of TGF-β pathway is decreased, and the activity of PI3K pathway is increased, and / or the activity of Notch pathway is decreased.The increase and / or decrease is preferably a monotonic increase and / or a monotonic decrease.The activity of TGF-β pathway is decreased, and the activity of PI3K pathway is increased, and the activity of Notch pathway is decreased, and / or the activity of Notch pathway is decreased, and / or the activity of TGF-β pathway is decreased, and / or the activity of Notch pathway is decreased, and / or the activity of TGF-β pathway is decreased, and / or the activity of Notch pathway is decreased, and / or the activity of TGF-β pathway is decreased, and / or the activity of Notch pathway is decreased, and / or the activity of TGF-β pathway is decreased, and / or the activity of Notch pathway is decreased, and / or the activity of TGF-β pathway is decreased, and / or the activity of TGF-β pathway is increased, and / or the activity of Notch pathway is decreased, and / or the activity of TGF-β ...
[0048] According to a preferred embodiment of the present invention and its various embodiments, the three or more TGF-β target genes are selected from the group consisting of ANGPTL4, CDC42EP3, CDKN1A, CTGF, GADD45A, GADD45B, HMGA2, ID1, IL11, JUNB, PDGFB, PTHLH, SERPINE1, SGK1, SKIL, SMAD4, SMAD5, SMAD6, SMAD7, SNAI2, VEGFA, more preferably from the group consisting of ANGPTL4, CDC42EP3, CDKN1A, CTGF, GADD45B, ID1, IL11, JUNB, SERPINE1, PDGFB, SKIL, SMAD7, SNAI2, and VEGFA, and most preferably from the group consisting of ANGPTL4, CDC42EP3, ID1, IL11, JUNB, SERPINE1, SKIL, and SMAD7.
[0049] According to a preferred embodiment of the present invention and its various embodiments, the three or more PI3K-FOXO target genes are selected from the group consisting of AGRP, BCL2L11, BCL6, BNIP3, BTG1, CAT, CAV1, CCND1, CCND2, CCNG2, CDKN1A, CDKN1B, ESR1, FASLG, FBXO32, GADD45A, INSR, MXI1, NOS3, PCK1, POMC, PPARGC1A, PRDX3, RBL2, SOD2, TNFSF10, preferably from the group consisting of FBXO32, BCL2L11, SOD2, TNFSF10, BCL6, BTG1, CCNG2, CDKN1B, BNIP3, GADD45A, INSR, and MXI1.
[0050] According to a preferred embodiment of the present invention and its various embodiments, the three or more Notch target genes are selected from the group consisting of CD28, CD44, DLGAP5, DTX1, EPHB3, FABP7, GFAP, GIMAP5, HES1, HES4, HES5, HES7, HEY1, HEY2, HEYL, KLF5, MYC, NFKB2, NOX1, NRARP, PBX1, PIN1, PLXND1, PTCRA, SOX9, and TNC, and preferably Two or more Notch target genes are selected from the group consisting of DTX1, HES1, HES4, HES5, HEY2, MYC, NRARP, and PTCRA, and one or more Notch target genes are selected from the group consisting of CD28, CD44, DLGAP5, EPHB3, FABP7, GFAP, GIMAP5, HES7, HEY1, HEYL, KLF5, NFKB2, NOX1, PBX1, PIN1, PLXND1, SOX9, and TNC.
[0051] In this regard, reference is made in particular to the sequence listings of the target genes provided above as follows: TGF-β: ANGPTL4, CDC42EP3, CDKNIA, CDKN2B, CTGF, GADD45A, GADD45B, HMGA2, ID1, IL11, SERPINE1, INPP5D, JUNB, MMP2, MMP9, NKX2- 5. OVOL1, PDGFB, PTHLH, SGK1, SKIL, SMAD4, SMAD5, SMAD6, SMAD7. SNAI1, SNAI2, TIMP1, and VEGFA(WO2016 / 062891, WO2016 / 062893); PI3K-FOXO:AGRP, BCL2L11, BCL6, BNIP3, BTG1, CAT, CAV1, CCND1, CCND 2. CCNG2, CDK1A, CDK1B, ESRl, FASLG, FBX032, GADD45A, INSR, MXIl, NOS 3. PCKl, POMC, PPARGCIA, PRDX3, RBL2, SOD2, and TNFSF10(WO2015 / 1016). 35);ATP8A1, BCL2L11, BNIP3, BTGl, ClOorflO, CAT, CBLB, CCNDl, CCND2 CDKNIB, DDB1, DYRK2, ERBB3, EREG, ESRl, EXT1, FASLG, FGFR2, GADD45 A, IGF1R, IGFBP1, IGFBP3, INSR, LGMN, MXI1, PPM1D, SEMA3C, SEPP1, SES N1, SLC5A3, SMAD4, SOD2, TLE4, and TNFSF10(WO2016 / 062892, WO2016 / 06 2893);SOD2, BNIP3, MXI1, PCK1, PPARGC1A, and CAT(EP16200697.7); Notch:CD28, CD44, DLGAP5, DTX1, EPHB3, FABP7, GFAP, GIMAP5, HES1, HES4, HES5, HES7, HEY1, HEY2 HEYL, KLF5, MYC, NFKB2, NOX1, NRARP, PBX1, PIN1, PLXND1, PTCRA, SOX9, and TNC(EP17194288.1).
[0052] The set of target genes found to best indicate the activity of the respective cell signaling pathway based on, for example, microarray / RNA sequencing-based studies using Bayesian or (pseudo)linear models can be converted into, for example, multiplex quantitative PCR assays or dedicated microarray biochips to be performed on samples of interest. The gene sequences described herein can be selected to select, for example, primer-probe sets or oligonucleotides for RT-PCR for the development of microarrays. In order to develop such FDA-approved tests for the determination of pathway activity and risk scores, standardized test kits need to be developed, which need to be clinically validated in clinical trials to obtain regulatory approval.
[0053] According to a third aspect, the present invention relates to an apparatus comprising a digital processor configured to perform the method of the first and / or second aspect of the invention and its various embodiments for determining an airway abnormality factor indicative of whether a subject has abnormal airway epithelium or a risk score indicative of the risk of a subject with abnormal airway epithelium developing airway cancer.Thus, the present invention relates to an apparatus comprising a digital processor configured to perform the method of the first and / or second aspect of the invention and its various embodiments.
[0054] According to a fourth aspect, the present invention relates to a non-transitory storage medium having stored thereon instructions executable by a digital processing device to implement the method of the first and / or second aspects of the present invention and its various embodiments for determining an airway abnormality factor indicative of whether a subject has abnormal airway epithelium or a risk score indicative of the risk of a subject with abnormal airway epithelium developing airway cancer. The non-transitory storage medium is a computer readable storage medium, such as a hard drive or other magnetic storage medium, an optical disk or other optical storage medium, a random access memory (RAM), a read-only memory (ROM), a flash memory, or other electronic storage medium, a network server, etc. The digital processing device is a handheld device (such as a personal data assistant or smartphone), a notebook computer, a desktop computer, a tablet computer or device, a remote network server, etc. Thus, the present invention relates to a non-transitory storage medium having stored thereon instructions executable by a digital processing device to implement the method of the first and / or second aspects of the present invention and its various embodiments.
[0055] According to a fifth aspect, the present invention relates to a computer program for determining an airway abnormality factor indicative of whether a subject has abnormal airway epithelium or has a risk score indicative of the risk of developing airway cancer, comprising program code means for causing a digital processing device to perform the method according to the first and / or second aspect of the invention and its various embodiments when said computer program is executed on said digital processing device. The computer program is stored / distributed on a suitable medium, for example an optical storage medium or a solid-phase medium provided together with or as part of other hardware, but also in other forms, for example via the internet or other wired or wireless telecommunication systems. Thus, the present invention relates to a computer program comprising program code means for causing a digital processing device to perform the method according to the first and / or second aspect of the invention and its various embodiments when said computer program is executed on said digital processing device.
[0056] According to a sixth aspect, the present invention relates to a kit for determining an airway abnormality factor indicative of whether a subject has abnormal airway epithelium or a risk score indicative of the risk of a subject with abnormal airway epithelium developing airway cancer, the kit comprising components for determining the expression levels of at least three target genes of the TGF-β cell signaling pathway, at least three target genes of the PI3K-FOXO cell signaling pathway, at least three target genes of the Notch cell signaling pathway, and / or at least three target genes of the HH cell signaling pathway.Preferably, the kit comprises components for determining the expression levels of at least three target genes of the TGF-β cell signaling pathway, at least three target genes of the PI3K-FOXO cell signaling pathway, and / or at least three target genes of the Notch cell signaling pathway. In a preferred embodiment of the kit according to the invention, the three or more TGF-β target genes are selected from the group consisting of ANGPTL4, CDC42EP3, CDKN1A, CTGF, GADD45A, GADD45B, HMGA2, ID1, IL11, JUNB, PDGFB, PTHLH, SERPINE1, SGK1, SKIL, SMAD4, SMAD5, SMAD6, SMAD7, SNAI2, and VEGFA, more preferably ANGPTL4, CDC42EP3, CDKN1A, CTGF, GAD or the three or more TGF-β target genes are selected from the group consisting of CDC42EP3, GADD45B, HMGA2, ID1, IL11, JUNB, SERPINE1, PDGFB, SKIL, SMAD7, SNAI2, and VEGFA, most preferably from the group consisting of ANGPTL4, CDC42EP3, ID1, IL11, JUNB, SERPINE1, SKIL, and SMAD7; or the three or more TGF-β target genes are selected from the group consisting of CDC42EP3, GADD45B, HMGA2, ID1, JUNB, OVAL1, VEGFA, SGK1; the three or more PI3K-FOXO target genes are selected from the group consisting of AGRP, BCL2L11, BCL6, BNIP3, BTG1, CAT, CAV1, CCND1, CCND2, CCNG2, CDKN1A, CDKN1B, ESR1, FASLG, FBXO32, GADD45A, INSR, MXI1, NOS3, PCK1, POMC, PPARGC1A, PRDX3, RBL2, SOD2, TNFSF10, preferably from the group consisting of FBXO32, BCL2L11, SOD2, TNFSF10, BCL6, BTG1, CCNG2, CDKN1B, BNIP3, GADD45A, INSR, and MXI1; The three or more Notch target genes are selected from the group consisting of CD28, CD44, DLGAP5, DTX1, EPHB3, FABP7, GFAP, GIMAP5, HES1, HES4, HES5, HES7, HEY1, HEY2, HEYL, KLF5, MYC, NFKB2, NOX1, NRARP, PBX1, PIN1, PLXND1, PTCRA, SOX9, and TNC, preferably two or more Notch target genes are selected from the group consisting of DTX1, HES1, HES4, HES5, HEY2, MYC, NRARP, and PTCRA, and one or more Notch target genes are selected from the group consisting of CD28, CD44, DLGAP5, EPHB3, FABP7, GFAP, GIMAP5, HES7, HEY1, HEYL, KLF5, NFKB2, NOX1, PBX1, PIN1, PLXND1, SOX9, and TN.
[0057] Another aspect of the present invention relates to the use of components for determining the expression levels of at least three target genes of the TGF-β cell signaling pathway, at least three target genes of the PI3K-FOXO cell signaling pathway, at least three target genes of the Notch cell signaling pathway, and / or at least three target genes of the HH cell signaling pathway for the manufacture of a kit for determining an airway abnormality factor indicative of whether a subject has abnormal airway epithelium or a risk score indicative of the risk of a subject with abnormal airway epithelium developing airway cancer.
[0058] The kit is in particular a quantitative kit, ie it allows the quantification of expression levels.
[0059] The kit comprises one or more components or means for measuring (in particular quantifying) the expression level of the target genes selected from the group consisting of DNA array chips, oligonucleotide array chips, protein array chips, antibodies, a plurality of probes, e.g. labeled probes, a set of RNA reverse transcriptase sequencing components, and / or DNA, including RNA or cDNA, amplification primers. In a preferred embodiment, the kit is selected from the group consisting of qPCR, multiplex qPCR, multiplex qPCR, ddPCR, RNAseq, RNA expression arrays, and mass spectrometry. In one embodiment, the kit comprises a set of labeled probes directed to a portion of the sequence of the mRNA or cDNA of the target genes described herein. In one embodiment, the kit comprises a set of primers and probes directed to a portion of the sequence of the mRNA or cDNA of the target genes. In one embodiment, the labeled probes are contained in a standardized 96-well plate. In one embodiment, the kit further comprises primers or probes directed to a set of reference genes. Such reference genes may be constitutively expressed genes useful in normalizing or standardizing the expression levels of the target gene expression levels, e.g. as described herein.
[0060] Thus, in one embodiment, the present invention relates to a kit comprising: components for determining the expression levels of at least three target genes of the TGF-β cell signaling pathway, at least three target genes of the PI3K-FOXO cell signaling pathway, and / or at least three target genes of the Notch cell signaling pathway; The three or more TGF-β target genes, e.g., 3, 4, 5, 6, 7, 8, 9, 10, 11, 12 or more genes, are selected from the group consisting of ANGPTL4, CDC42EP3, CDKN1A, CTGF, GADD45A, GADD45B, HMGA2, ID1, IL11, JUNB, PDGFB, PTHLH, SERPINE1, SGK1, SKIL, SMAD4, SMAD5, SMAD6, SMAD7, SNAI2, VEGFA, more preferably the three or more TGF-β target genes, e.g., 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13 or 14 genes, are selected from the group consisting of ANGPTL4, CDC42EP3, CDKN1A, CTGF, GF, GADD45B, ID1, IL11, JUNB, SERPINE1, PDGFB, SKIL, SMAD7, SNAI2, and VEGFA, and most preferably, three or more TGF-β target genes, e.g., three, four, five, six, seven, or eight genes, are selected from the group consisting of ANGPTL4, CDC42EP3, ID1, IL11, JUNB, SERPINE1, SKIL, and SMAD7; or three or more TGF-β target genes, e.g., three, four, five, six, seven, or eight genes, are selected from the group consisting of CDC42EP3, GADD45B, HMGA2, ID1, JUNB, OVAL1, VEGFA, SGK1; and / or Three or more PI3K-FOXO target genes, for example, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12 or more genes, are AGRP, BCL2L11, BCL6, BNIP3, BTG1, CAT, CAV1, CCND1, CCND2, CCNG2, CDKN1A, CDKN1B, ESR1, FASLG, FBXO32, GADD45A, INSR, MXI1, NOS3, PCK1, POMC, PPARGC 1A, PRDX3, RBL2, SOD2, TNFSF10, preferably three or more PI3K-FOXO genes, e.g., 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, or 13 genes, are selected from the group consisting of FBXO32, BCL2L11, SOD2, TNFSF10, BCL6, BTG1, CCNG2, CDKN1B, BNIP3, GADD45A, INSR, and MXI1; and / or The three or more Notch target genes, e.g., 3, 4, 5, 6, 7, 8, 9, 10, 11, 12 or more genes, are selected from the group consisting of CD28, CD44, DLGAP5, DTX1, EPHB3, FABP7, GFAP, GIMAP5, HES1, HES4, HES5, HES7, HEY1, HEY2, HEYL, KLF5, MYC, NFKB2, NOX1, NRARP, PBX1, PIN1, PLXND1, PTCRA, SOX9, and TNC, and preferably, two or more Notch target genes, e.g., 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12 or more genes, are selected from the group consisting of CD28, CD44, DLGAP5, DTX1, EPHB3, FABP7, GFAP, GIMAP5, HES1, HES4, HES5, HES7, HEY1, HEY2, HEYL, KLF5, MYC, NFKB2, NOX1, NRARP, PBX1, PIN1, PLXND1, PTCRA, SOX9, and TNC. The 6, 7 or 8 genes are selected from the group consisting of DTX1, HES1, HES4, HES5, HEY2, MYC, NRARP, and PTCRA, and one or more Notch target genes, such as 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12 or more genes are selected from the group consisting of CD28, CD44, DLGAP5, EPHB3, FABP7, GFAP, GIMAP5, HES7, HEY1, HEYL, KLF5, NFKB2, NOX1, PBX1, PIN1, PLXND1, SOX9, and TNC. Preferably, the kit is for or suitable for determining an airway abnormality factor indicative of whether a subject has abnormal airway epithelium or has a risk score indicative of a risk of developing airway cancer for a subject having abnormal airway epithelium.
[0061] In some embodiments, the kit is not a microarray for determining the expression level of several thousand target genes.For example, the kit of the present invention includes the components for determining the expression level of not more than 1000 target genes, not more than 700 target genes, not more than 500 target genes, not more than 200 target genes, not more than 100 target genes, in addition to the components required for the specific target genes disclosed herein.
[0062] The kit further comprises the apparatus of the third aspect, the non-transitory storage medium of the fourth aspect, or the computer program of the fifth aspect.
[0063] According to a seventh aspect, the present invention relates to a kit for use in a method for diagnosing or prognosing whether a subject has abnormal airway epithelium or whether a subject with abnormal airway epithelium will develop airway cancer, the kit comprising components for determining the expression levels of at least three target genes of the TGF-β cell signaling pathway, at least three target genes of the PI3K-FOXO cell signaling pathway, at least three target genes of the Notch cell signaling pathway and / or at least three target genes of the HH cell signaling pathway. According to a preferred embodiment, the method of diagnosis or prognosis comprises the steps of extracting an epithelial cell sample from the airway of the subject and subjecting the extracted epithelial cell sample to a method according to the first and / or second aspect of the invention.
[0064] Another aspect of the invention is a method of diagnosing or prognosing whether a subject has abnormal airway epithelium or whether a subject with abnormal airway epithelium will develop airway cancer, the method comprising the steps of providing an epithelial cell sample of the airway of the subject or extracting an epithelial cell sample from the airway of the subject and subjecting the prepared or extracted epithelial cell sample to a method according to the first and / or second aspects of the invention.
[0065] Therefore, the present invention further relates to a method for diagnosing or prognosing whether a subject has abnormal airway epithelium or whether a subject with abnormal airway epithelium will develop airway cancer, in vivo or ex vitro, using a kit, the kit comprising components for determining the expression levels of at least three target genes of the TGF-β cell signaling pathway, at least three target genes of the PI3K-FOXO cell signaling pathway, at least three target genes of the Notch cell signaling pathway. Preferably, in said method for diagnosis or prognosis in vivo or ex vitro, an epithelial cell sample of the airway of the subject is prepared or an extracted epithelial cell sample from the airway of the subject is prepared, and the prepared or extracted epithelial cell sample is subjected to the method according to the first and / or second aspect of the present invention. Preferably, said method further comprises the step of preparing an epithelial cell sample of the airway of the subject or extracting an epithelial cell sample from the airway of the subject.
[0066] Preferably, in said method for in vivo or ex vitro diagnosis or prognosis, said component for determining the expression levels of at least three target genes of the TGF-β cell signaling pathway, at least three target genes of the PI3K-FOXO cell signaling pathway, at least three target genes of the Notch cell signaling pathway comprises three or more TGF-β target genes, for example 3, 4, 5, 6, 7, 8, 9, 10, 11, 12 or more genes are selected from the group consisting of ANGPTL4, CDC42EP3, CDKN1A, CTGF, GADD45A, GADD45B, HGF, GADD45C, GADD45D, GADD45E, GADD45F, GADD45G, GADD45H, GADD45I ... and more preferably, the three or more TGF-β target genes, e.g., 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, or 14 genes, are selected from the group consisting of ANGPTL4, CDC42EP3, CDKN1A, CTGF, GADD45B, ID1, IL11, JUNB, SERPINE1, PDGFB, SKIL, SMAD7, SNAI2, and VEGFA. and most preferably, three or more TGF-β target genes, e.g., three, four, five, six, seven, or eight genes, are selected from the group consisting of ANGPTL4, CDC42EP3, ID1, IL11, JUNB, SERPINE1, SKIL, and SMAD7; or three or more TGF-β target genes, e.g., three, four, five, six, seven, or eight genes, are selected from the group consisting of CDC42EP3, GADD45B, HMGA2, ID1, JUNB, OVAL1, VEGFA, SGK1, and three or more PI3K-FOXO targets. The genes, e.g., 3, 4, 5, 6, 7, 8, 9, 10, 11, 12 or more genes, are selected from the group consisting of AGRP, BCL2L11, BCL6, BNIP3, BTG1, CAT, CAV1, CCND1, CCND2, CCNG2, CDKN1A, CDKN1B, ESR1, FASLG, FBXO32, GADD45A, INSR, MXI1, NOS3, PCK1, POMC, PPARGC1A, PRDX3, RBL2, SOD2, TNFSF10, and preferably, three or more PI3K-FOXO genes, e.g., 3, 4, 5, 6, 7, 8, 9, 10, 11, 12 or more genes.the 9, 10, 11, 12, or 13 genes are selected from the group consisting of FBXO32, BCL2L11, SOD2, TNFSF10, BCL6, BTG1, CCNG2, CDKN1B, BNIP3, GADD45A, INSR, and MXI1; and three or more Notch target genes, for example, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, or more genes are selected from the group consisting of CD28, CD44, DLGAP5, DTX1, EPHB3, FABP7, GFAP, GIMAP5, HES1, HES4, HES5, HES7, HEY1, HEY2, HEYL, KLF5, MYC, NFKB2, NOX1, NRARP, PBX1, PIN1, PLXND1, PTCRA, Preferably, two or more Notch target genes, e.g., 2, 3, 4, 5, 6, 7, or 8 genes, are selected from the group consisting of DTX1, HES1, HES4, HES5, HEY2, MYC, NRARP, and PTCRA, and one or more Notch target genes, e.g., 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, or more genes, are selected from the group consisting of CD28, CD44, DLGAP5, EPHB3, FABP7, GFAP, GIMAP5, HES7, HEY1, HEYL, KLF5, NFKB2, NOX1, PBX1, PIN1, PLXND1, SOX9, and TNC. Preferably, the kit is as described in the kit according to the sixth aspect of the present invention.
[0067] In a further embodiment, the method of diagnosing or prognosing in vivo or ex vitro whether a subject has abnormal airway epithelium or whether a subject with abnormal airway epithelium will develop airway cancer according to the present invention further comprises determining an airway abnormality factor indicative of whether the subject has abnormal airway epithelium based on a combination of cell signaling pathway activity in an epithelial cell sample derived from the subject's airway, the cell signaling pathway comprising two or more cell signaling pathways selected from the group consisting of the TGF-β pathway, the PI3K-FOXO pathway, and the Notch pathway, and the determining the signaling pathway abnormality factor is further based on a reference activity of each cell signaling pathway, the reference activity reflecting the activity of each cell signaling pathway found in the airway epithelium of a healthy subject.
[0068] According to an eighth aspect, the present invention relates to a kit comprising parts for determining the expression levels of at least three target genes of the TGF-β cell signaling pathway, at least three target genes of the PI3K-FOXO cell signaling pathway, at least three target genes of the Notch cell signaling pathway. Preferably, the kit is for or suitable for determining a risk score indicative of whether a subject has abnormal airway epithelium or whether a subject with abnormal airway epithelium will develop airway cancer.
[0069] In a further preferred embodiment, the invention relates to the use of a kit according to the sixth or eighth aspect of the invention in determining an airway abnormality score indicating that a subject has abnormal airway epithelium or whether a subject with abnormal airway epithelium will develop airway cancer or for determining a risk score.
[0070] One advantage of the present invention resides in a clinical decision support (CDS) system adapted to provide clinical recommendations by determining a treatment for a subject based on a combination of pathway activities described herein, e.g., as indicated by an airway abnormality factor and / or a risk score determined based on a combination of pathway activities.
[0071] Another advantage resides in the CDS system being adapted to assign a subject to at least one of a plurality of risk groups with varying risks of the subject developing airway cancer as indicated by an airway abnormality factor and / or risk score determined based on a combination of activity of two or more cell signaling pathways described herein.
[0072] Another advantage resides in combining the risk score, indicative of a subject's risk for developing airway cancer and determined based on a combination of the activity of two or more cell signaling pathways described herein, with one or more additional risk scores obtained from one or more additional predictive tests.
[0073] The invention described herein includes, for example, Predicting whether the airway epithelium is pre-malignant; and / or predicting whether a person is at high risk of developing respiratory tract cancer (as defined herein); and / or Predicting whether a person is at high risk of developing lung cancer; and / or Predicting whether a person is at high risk of developing squamous cell lung cancer; and / or Predicting whether a person is at high risk of developing lung adenocarcinoma; and / or predicting whether a person would benefit from localized treatment to prevent the development of cancer; and / or The patient has lung cancer and / or prognosis and / or prediction based on the combined activity of two or more cell signaling pathways, and / or Predicting the efficacy of a drug, e.g. chemotherapy and / or hormonal therapy, based on the combined activity of two or more cell signaling pathways, and / or monitoring the efficacy of a drug, based on the combined activity of two or more cell signaling pathways, and / or Determining the frequency of monitoring based on a combination of the activities of two or more cell signaling pathways or, more particularly, the frequency of monitoring therapeutic response; and / or The development of drugs based on the combined activity of two or more cell signaling pathways, and / or Development of an assay based on the combination of activities of two or more cell signaling pathways; and / or Predicting whether a person is at risk of developing invasive airway cancer; and / or Predicting whether a person is at risk of disease progression, and / or predicting or diagnosing whether a person has a reduced risk following treatment (e.g., chemoprevention), and / or Predicting whether a nodule or abnormality detected by an imaging procedure such as a CT scan is likely to be malignant or benign; and / or Cancer staging based on the combined activity of two or more cell signaling pathways in each case the cell signaling pathway comprises the TGF-β pathway, the PI3K-FOXO pathway, and / or the Notch pathway.
[0074] Further advantages will become apparent to those of ordinary skill in the art upon reading and understanding the accompanying drawings, the following description, and especially upon reading the detailed examples provided herein below.
[0075] This application has described certain preferred embodiments. Modifications and alterations will occur to others upon reading and understanding the above detailed description. It is intended that this application be construed as including all such modifications and alterations insofar as they come within the scope of the appended claims or the equivalents thereof.
[0076] Other variations to the disclosed embodiments can be understood and effected by those skilled in the art in practicing the claimed invention, from a study of the drawings, the disclosure, and the appended claims.
[0077] It is to be understood that the method of the first and second aspects, the apparatus of the third method, the non-transitory storage medium of the fourth method, the computer program of the fifth aspect and the kits of the sixth, seventh and eighth aspects have similar and / or identical preferred embodiments, in particular as defined in the dependent claims.
[0078] In the claims, the word "comprising" does not exclude other elements or steps, and the indefinite article "a" or "an" does not exclude a plurality.
[0079] A single unit or device fulfills the functions of several items recited in the claims. The mere fact that certain measures are recited in mutually different dependent claims does not indicate that a combination of these measures cannot be used to advantage.
[0080] Calculations such as the determination of a risk score performed by one or several units or devices can be performed by any other number of units or devices.
[0081] It shall be understood that a preferred embodiment of the invention may also be any combination of the dependent claims or the above embodiments with the respective independent claim.
[0082] These and other aspects of the invention will be apparent from and elucidated with reference to the embodiments described hereinafter. [Brief description of the drawings]
[0083] [Figure 1] 1A-1C show exemplary results of the determination of signaling pathway activity measured in normal small airway epithelial cells obtained by scraping from non-smokers (n=13) (left box, labeled "NS") and heavy smokers (n=18) (right box, labeled "S") (FIG. 1A: TGF-β, FIG. 1B: Notch, FIG. 1C: PI3K-FOXO). Log2 odds pathway scores are shown. [Diagram 2]FIG. 2 shows an exemplary calculation of thresholds for abnormal pathway activity in airway epithelial cells using TGF-β pathway activity using dataset GSE10006. Here, pathway activity scores for each sample in the dataset are shown separately for "non-smokers" and "smokers". The mean pathway activity scores are calculated for both groups, as well as the standard deviation (SD) (shown in FIG. 2 as small rectangles). The threshold for abnormal low TGF-β activity was set as the mean pathway activity score minus 1 SD or the mean minus 2 SD of the TGF-β pathway scores of normal small airways in healthy non-smokers, shown in FIG. 2 as horizontal dotted lines labeled "1 SD" and "2 SD", respectively. If 1 SD is taken as the threshold, TGF-β pathway scores below this level occur in only 15.8% of healthy non-smokers, and if 2 SD is taken as the lower threshold, it occurs in only 2.2% of the population. Values below this threshold (depending on the 1 SD or 2 SD threshold) are considered to be associated with increased risk from abnormal airway epithelium. For the development of the centroid computer method described here, sample pathway analysis results below a defined 2SD threshold are considered abnormal. In this case, the abnormal pathway activity is characterized by a loss of tumor-suppressing TGF-β activity. In the smoker group of this data set, it is expected that only a subgroup, but not all, of the smokers have abnormal airway epithelium. This subgroup can be identified by applying the TGF-β pathway activity threshold of normal epithelium to this group. Indeed, a relatively large number of samples had TGF-β pathway activity below the 1SD and 2SD thresholds, respectively (Figure 2B). For the subsequent calibration of the centroid computer model, samples with TGF-β pathway scores below 2SD of normal are considered as the calibration set for abnormal samples. Meanwhile, non-smoker samples with TGF-β pathway scores higher than the 2SD threshold constitute the calibration set for normal samples. For the linear model, the 1SD threshold is taken as an example threshold below which the pathway activity score is considered abnormally low. This can be done in a corresponding manner for other signaling pathways. Figure 2B further shows the calibration data in the form of boxes. The left square shows the calibration set for normal samples "NS". The right square shows the calibration set for abnormal samples "AS". [Diagram 3] FIG. 1 is a schematic diagram of a clinical decision support (CDS) system configured to determine a risk score indicative of a subject's risk of developing airway cancer, as disclosed herein. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
[0084] The following embodiments merely illustrate particularly preferred methods and selected aspects related thereto. The teachings provided herein can be used to construct several tests and / or kits. The following examples should not be construed as limiting the scope of the invention.
[0085] The present invention relates to a method that can identify individuals, particularly smokers, who have abnormal / pre-malignant changes in the airway epithelium and / or are at increased risk of developing lung cancer. Measurement of signal transduction pathway activity can identify abnormal signal transduction pathway activity indicative of abnormal airway epithelium and increased likelihood of pre-malignant changes. According to the present invention, the activity of two or more pathways is determined and converted into an Airway Abnormality Factor in the (S / L) APAS, also referred to as the (Small / Large) Airway Pathway Abnormality Score, which identifies individuals with abnormal airway epithelium potentially at increased risk of developing lung cancer.
[0086] The present invention provides a risk score and / or airway epithelial abnormality factor / score indicative of the presence of pre-malignant changes in epithelial cells from the airways based on the combined activity of the PI3K-FOXO pathway, Notch, HH, and / or TGF-β pathway, and interpretation of pathway results using a computer model that provides the risk score and / or airway epithelial abnormality factor. Analysis of epithelial cells from large or small airways can be used to detect early abnormalities that indicate a high risk of developing lung cancer. The risk score and / or airway epithelial abnormality factor can also be used to stratify patients for close monitoring or aid in the selection of treatments directed at restoring these pathways to their non-pathological state and reversing the pre-malignant changes.
[0087] Lung cancer develops in the epithelium of the large airways (bronchial branches distant from the trachea) or small airways. Using techniques such as airway scraping, nasal cell sampling (nose swabs), or bronchoalveolar lavage, epithelial cells can be obtained from the upper and lower airways in a relatively non-invasive manner for molecular analysis. This provides a possible means of identifying individuals, particularly smokers, who have these early proliferative changes in their airway epithelium and are at increased risk for developing lung cancer.
[0088] Signal transduction pathway analysis allows quantitative measurement of signal transduction pathway activity in epithelial cells obtained from the upper or lower airways, for example by bronchoalveolar lavage, scraping, or biopsy, based on inferring the activity of signal transduction pathways from measuring the mRNA levels of confirmed direct target genes of transcription factors associated with the respective signal transduction pathways (see, for example, W Verhaegh et al., 2014, supra; W Verhaegh, A van de Stolpe, 2014, supra). The steps of determining the activity of one or more pathways, combinations of multiple pathway activities, and their applications are carried out as described above. Models have been biologically validated for ER, AR, PI3K-FOXO, HH, Notch, TGF-β, Wnt, NFkB, and STAT1 / 2 and STAT3 pathways in several cell types, including epithelial cells.
[0089] The present invention focuses on the TGF-β pathway, the PI3K-FOXO pathway, the Notch pathway, and / or the HH pathway.
[0090] TGF-β is involved in cell proliferation, differentiation, immune cell activity, cell microenvironment, and other cellular processes. In normal and premalignant cells, TGF-β exerts tumor suppressive functions. However, in the progression to cancer cells, the tumor suppressive effect can be lost due to receptor inactivating mutations or selective loss of the inhibitory arm of the pathway (Massague et al., Cell, 2008, 134(2):215-230). Tobacco exposure can reduce TGF-β-mediated growth inhibition and apoptosis, indicating that smoking promotes tumorigenicity (Samanta et al., Cancer Prev Res, 2012, 5(3):453-63).
[0091] The PI3K-FOXO pathway is commonly hyperactivated in various types of cancer. Although tumors are likely sensitive to PI3K-FOXO pathway inhibitors, there is no reliable diagnostic test to assess functional PI3K-FOXO activity. Because the PI3K-FOXO pathway negatively regulates tumor-suppressive FOXO transcription factors, expression of FOXO target genes inversely correlates with PI3K activity (assuming no oxidative stress) (van Ooijen, 2018, supra). Lack of FOXO3 has been shown to result in increased susceptibility to smoking-induced inflammation, increased air spaces, and COPD (Hwang et al., J Immunol, 2011, 187(2):987-998). FOXO3 levels are significantly reduced in the lungs of smokers and patients with chronic obstructive pulmonary disease (COPD).
[0092] Notch signaling is involved in the control of cell proliferation, differentiation, and apoptosis. The tumor suppressive and tumor promoting functions of the different isoforms of Notch depend on the cellular and environmental context. For example, in some tumors, it has been described that Notch1 has a tumorigenic role in promoting tumor initiation, whereas Notch2 has a tumor suppressive role (Zou et al., Oncology Letters, 2018, 15:3415-3421). However, this can differ in different tumor types, for example in bladder cancer both NOTCH1 and NOTCH2 are tumor suppressive (Cancer Discov, 2014, 4(11):1252). The expression levels of the different isoforms differ among various histological subtypes of lung cancer (Chen et al., Journal of Cancer, 2017, 8(7):1292-1300). In one type of lung cancer, SCLC, the tumor suppressor function of the Notch pathway is lost, which even provides a therapeutic target, and drugs have been developed to increase Notch pathway activity in patients with SCLC (Nat Rev Clin Oncol, 2017, 14(9):549-561). The Notch pathway is downregulated in the airway epithelium of healthy smokers and smokers with chronic obstructive pulmonary disease, implying that this pathway is important in repairing smoking-induced injury (Tilley et al., Am J Respir Crit Care Med, 2009, 179(6):457-66).
[0093] Signaling pathway analysis of the Affymetrix U133 Plus 2.0 expression microarray dataset (GEO, a public dataset) with data from smokers and non-smokers We identified publicly available clinical studies in the GEO dataset database ( https: / / www.ncbi.nlm.nih.gov / gds / ) for Affymetrix Plus 2.0 that contained Affymetrix data from epithelial airway cells of non-smokers, and light and heavy cigarette smokers.
[0094] Dataset GSE10006. The groups studied were smokers' large airways (n=9), non-smokers' large airways (n=20), smokers' small airways (n=13), non-smokers' small airways (n=18), and COPD patient data as an independent dataset for validation purposes. The results of the analysis of TGF-beta, NOTCH pathway, and FOXO transcription factor activity are shown in Figure 1A-1C for normal small airway epithelium of non-smokers and smokers. The expression levels of target genes disclosed herein were collected from the dataset for each of the TGF-beta, PI3K-FOXO, and Notch pathways. Signaling pathway activity was then determined as described herein and compared between non-smokers and smokers. Pathway activity is displayed on a log2 odds scale and significant differences (Rank Wilcoxon test) are displayed.
[0095] Using this approach, it was found that the TGF-β and Notch pathways are lost in a subpopulation of heavy smokers (see Figure 1A and Figure 1B), while FOXO transcription factor activity is lost in a subgroup of smokers with activated PI3K growth factor pathways (see Figure 1C). This means that there is a distinctive pathway activity profile in the airway epithelial cells of a subgroup of heavy smokers compared to non-smokers. The loss of the tumor suppressor effect of TGF-β and Notch pathway activity is accompanied by increased activity of the proliferative PI3K pathway. This indicates abnormal proliferation and loss of tumor suppressor activity of Notch and TGF-β pathways in a subgroup of heavy smoking populations. Subjects in this subgroup are likely at high risk for developing some form of lung cancer. If these abnormalities are found in cells from the upper respiratory tract, they are likely at high risk for developing cancers such as squamous cell lung cancer, which typically occurs here, and if present in lower respiratory tract epithelial cells, the subjects are likely at high risk for developing adenocarcinomas, which typically occur in the lower respiratory tract.
[0096] It was also determined that epithelial cells from the trachea (upper respiratory tract) exhibited similar pathway activity abnormalities as found in lower respiratory tract epithelial cells and therefore could be used as a surrogate sample to provide an epithelial abnormality score and a cancer risk score (data not shown). Epithelial cells from the trachea are easily obtained and sampling is minimally invasive.
[0097] Using the dataset GSE19722, it was further found that the present invention is also applicable to stem / progenitor cells of epithelial cells (Dataset GSE19722.REF:Shaykhiev R,Wang R,Zwick RK,Hackett NR et al. Airway basal cells of healthy smokers express an embryonic stem cell signature relevant to lung cancer.Stem Cells 2013 Sep;31(9):1992-2002). Thus, in particular, cultured basal cells (BC), which constitute the stem / progenitor cells required for the regeneration of damaged epithelium, can be used as a surrogate for primary non-cultured epithelial cell samples. Reduced FOXO transcription factor activity in basal cells from smokers compared to non-smokers is an indication of increased PI3K pathway activity (see Table 1).
[0098] Table 1. Epithelial stem / progenitor cell analysis using dataset GSE19722. Small airway epithelial cells were harvested using a flexible bronchoscope and cultured. RNA was extracted one week later for Affymetrix Microarray analysis. Log2 odds values for FOXO activity are shown.
[0099] [Table 1]
[0100] The earliest damage to the airway epithelium associated with smoking is reflected by BC hyperplasia caused by increased activity of the PI3K pathway, as identified by the Philips pathway analysis described herein. In this case, BC were obtained by bronchoscopy from non-smokers and smokers and cultured on collagen for 1 week before analysis. Under these conditions, PI3K-FOXO activity was increased (not significantly) in cultured BC from smokers, indicating increased PI3K-FOXO pathway activity (and decreased FOXO activity) (see Table 1). Aberrant PI3K pathway activity in this cell type correlated with aberrant pathway activity in epithelial cell types, suggesting that the PI3K pathway activity seen in epithelial cells is of basal cell origin. This experiment indicates that BC-derived samples could in principle also be used to measure epithelial cell abnormalities.
[0101] Definition of normal pathway activity Reference pathway activity was defined for normal small airway epithelium based on pathway activity in healthy non-smoking subjects (GSE10006). Pathway activity measured in patient samples was considered abnormal if the measured pathway activity was more than one standard deviation (>1SD) or more than two standard deviations (>2SD) lower than the mean of normal pathway activity for Notch, TGF-β pathway, and FOXO transcription factor activity. The determined means and standard deviations for each pathway are shown in Table 2.
[0102] Table 2. Mean and standard deviation of TGF-β, NOTCH pathway, and FOXO transcription factor activity
[0103] [Table 2]
[0104] Based on these values, a lower threshold of pathway activity was calculated below which pathway activity was considered abnormal in airway epithelial cells. Lower threshold for TGF-β activity: −15.17 (1 SD) and −17.12 (2 SD); Lower threshold of NOTCH activity: 8.79 (1 SD) and 6.57 (2 SD); Lower threshold of PI3K-FOXO activity: -0.14 (1 SD) and -2.40 (2 SD) The lower threshold for TGF-β pathway activity calculated in this manner is shown as a horizontal dotted line in Figures 2A and 2B.
[0105] Computer model development and calibration Due to the fact that individual pathway activity shows variability between patient samples, a computer model was advantageously employed to interpret multiple pathway activities and provide a probability that the analyzed epithelial cell sample is not normal. This is believed to provide an indication of a pre-malignant state. For model calibration, Affymetrix U133Plus2.0 data of healthy smokers and non-smokers small airway epithelium from the dataset GSE10006 was used. For validation purposes from the same dataset, independent data from a sample of a smoker with COPD was used.
[0106] The model may be a linear, centroid, Bayesian model, or another model as described herein. Models may be developed separately for lower and upper respiratory tract epithelial cells, which may provide abnormality scores indicative of risk in developing various lung cancer types.
[0107] In the present example, centroid and linear models were used, but those skilled in the art will appreciate that Bayesian and other models could be used as well.
[0108] Centroid Model An exemplary computer model was developed that uses pathway activity of PI3K, TGF-β, and Notch measured in epithelial cells obtained from the lower (small) airways using scrapings as input to calculate a centroid model (small / large) airway pathway abnormality score.
[0109] The selection of samples for calibrating the centroid computer model was as follows (see FIG. 2B).
[0110] Selection of "normal samples" based on healthy / normal small airways (non-smokers) exceeding a 2×SD threshold of non-smoker data.
[0111] Selection of "abnormal samples" was based on healthy / normal small airways (smokers) below a threshold of 2×SD of the nonsmoker data.
[0112] Models were calibrated using a combination of TGF-β and FOXO activity (see Table 3) or a combination of TGF-β, FOXO, and Notch activity (see Table 4). Prior to using the microarray data, extensive quality control (QC) was performed on the Affymetrix data from each individual sample as described elsewhere (A van de Stolpe et al., 2019, supra). Only samples that passed QC were used for further analysis.
[0113] Table 3. Calibration data for models based on the combination of TGF-β, NOTCH, and FOXO activity using the GSE10006 dataset. Each row represents an individual sample. Sample "NS" represents a sample from a non-smoker (calibration normal sample). Sample "AS" represents a sample from a smoker (calibration abnormal sample). All samples passed quality control.
[0114] [Table 3]
[0115] Table 4. Calibration data for models based on the combination of TGF-β and FOXO activity using the GSE10006 dataset. Each row represents an individual sample. Sample "NS" represents a sample from a non-smoker (calibration normal sample). Sample "AS" represents a sample from a smoker (calibration abnormal sample). All samples passed quality control.
[0116] [Table 4]
[0117] Smokers with COPD from the same dataset were used to validate the model. Results are further shown in Tables 5-7 below.
[0118] Linear Models Another computer model was developed that provides a linear (small / large) airway pathway abnormality score. In this example, a score of 1 point was assigned to each abnormal pathway activity. Pathway activity was considered abnormal if the measured TGF-β or Notch pathway or FOXO activity was less than the mean -1SD or -2SD of normal non-smokers (see Table 2 above and Figure 2A for calculation of these thresholds). Otherwise, the respective pathway was assigned a score of "0". The scores were tallied to indicate the likelihood that the airway epithelium was abnormal and the patient was probably at high risk for developing lung cancer. The higher the score, the more likely the epithelium was abnormal and at risk for developing lung cancer. The score is called the APAS score (Airway Pathway Activity Score). A score of 0 is normal (low risk) and a score of 3 indicates the most abnormal epithelium and is predicted to confer the highest risk for developing lung cancer.
[0119] Validation of computer models For validation purposes, independent data on small airway epithelium from the GSE10006 dataset of smokers with early chronic obstructive pulmonary disease (COPD) or long-term COPD were then used to score abnormal epithelial status in smokers.
[0120] Calculation of the Airway Epithelial Aberration Score (APAS) using a linear computer model In this experiment, thresholds were calculated based on the mean and variance of each pathway activity measured in epithelial cells of the small airways of healthy non-smokers ("normal" airway epithelium). More specifically, the mean and standard deviation of TGF-β, Notch, and FOXO activity in normal epithelial cells were determined by the pathway analysis described herein. These values, along with the resulting 1SD and 2SD thresholds, are summarized in Table 2 and the section following this table and shown in Figure 2A.
[0121] As exemplarily shown for the TGF-β pathway using dataset GSE10006 (see Figures 2A and 2B), the threshold was defined as the mean pathway activity minus 2SD. The threshold indicates the transition from normal to abnormally low TGF-β activity. The 2SD threshold is shown for small airway epithelium. If the pathway activity was determined to be below the defined threshold / horizon, the sample was considered to have abnormal pathway activity. It is seen that a subgroup of smokers has abnormal TGF-β pathway activity, especially in the small airways (see Figure 2B). In this case, the abnormal pathway activity is characterized by a loss of tumor suppressive TGF-β pathway activity. This group of subjects is presumed to be at increased risk of developing airway cancer.
[0122] For linear models, a pathway aberration factor was assigned for each assessed pathway activity based on these thresholds. If the pathway activity was judged to be aberrant, a factor of "1" was assigned, otherwise a factor of "0" was assigned. For example, if the activity of the TGF-β pathway was judged to be below this threshold, a score of 1 point was assigned to the respective pathway. A corresponding notion was applied for FOXO activity. The pathway aberration factors were then summed to obtain the APAS score.
[0123] For validation purposes using a linear model, deviations from normal pathway activity in small airway epithelium were evaluated and an APAS score was determined based on a combination of TGF-β pathway activity and FOXO transcription factor activity. An APAS score of "0" represents normal airway epithelium (low risk). A higher score indicates a higher probability that the airway epithelium is abnormal. All samples used for validation passed the QC described herein. The results of the validation are shown in Table 5 along with the calibration data.
[0124] Table 5. Calibration and validation results of the linear model based on the combined activity of TGF-β and FOXO. The calibration results correspond to the results shown in Table 4. The validation was performed with samples from smokers with early (sample "eCOPD") and established COPD (sample "COPD") from the GSE10006 dataset. All samples passed the QC described herein.
[0125] [Table 5] TIFF0007583999000006.tif209170
[0126] Using the linear model, the incidence of individuals with abnormal airway epithelium was greatest in smokers with established COPD and less in those with early COPD, which is consistent with the expected overall risk of lung cancer in COPD patients. Thus, the model performed as expected.
[0127] Validation of the centroid model The model was validated using smokers with early and established COPD from the same dataset (GSE10006). The model calculates a distance score of pathway activity found in the sample to clusters of normal / healthy pathway activity and abnormal epithelial pathway activity. Based on the calibration of the model, this score defines whether the analyzed sample is considered normal small airway epithelium or abnormal. The model can be used with a combination of TGF-β and FOXO activity (see Table 6) or a combination of TGF-β, FOXO, and Notch activity (see Table 7).
[0128] Table 6. Validation data for the centroid model based on combined FOXO and TGF-β activity using samples from smokers with early and established COPD from the GSE10006 dataset. All samples passed the QC described herein.
[0129] [Table 6]
[0130] Table 7. Validation data for the centroid model based on combined TGF-β, NOTCH, and FOXO activity using samples from smokers with early and established COPD from the GSE10006 dataset. All samples passed the QC described herein.
[0131] [Table 7]
[0132] It was seen that the model scored some early COPD samples as normal but consistently scored established / long-term COPD samples as abnormal, which is consistent with the expected risk of lung cancer in these two groups.
[0133] Clinical Use of the Method Apparently, not all (heavy) smokers develop lung cancer, which is why airway abnormality factors and / or risk scores are only detected in a subgroup of smokers. This variation among smokers also indicates a need for the means and methods of the present invention that allow the identification of subjects with abnormal airway epithelium, in particular smokers.
[0134] The present invention is expected to reduce unnecessary invasive diagnostic procedures in heavy smokers and allow for earlier minimally invasive treatment of high-risk patients.
[0135] Furthermore, pathway analysis in airway epithelial cells could provide an additional diagnostic test for lung cancer and allow for the selection of systemic (targeted) therapies. 1. The described method can be used to screen smokers for pre-malignant changes that have a high risk of developing lung cancer, avoiding repeated exposure to radiation associated with imaging techniques such as CT scans for screening.Epithelial cells can be obtained from airway epithelium, and determination of TGF-β / Notch / PI3K-FOXO pathway activity combination or at least one of these pathway activities, preferably Notch and TGF-β, can be performed.The results can be interpreted in the screening model to predict the risk of pre-malignant changes being present, indicating a high risk of developing lung cancer associated with smoking. 2. If a smoker presents with symptoms (e.g. cough or shortness of breath) or abnormal imaging findings, epithelial cells can be obtained from the airway epithelium, e.g. using BAL, and a combined determination of TGF-β / Notch / PI3K-FOXO pathway activity can be performed and the results interpreted in a screening model to predict the risk of the presence of pre-malignant changes and indicate an increased risk of developing lung cancer associated with smoking. 3. If a lung lesion is detected on imaging such as a chest x-ray or CT scan and it is unclear whether this is a malignant lesion, performing pathway analysis on BAL cells can provide complementary information about the nature of the lesion with regard to its malignant potential, especially in smokers. 4. Classification based on pathway activity in diagnostic procedures, e.g., bronchoscopy readouts, can be aggregated to improve the sensitivity / specificity of methods assessing the probability of lung cancer. 5. In any of the above scenarios, the calculated risk score can be used to stratify patients for close monitoring and / or chemoprophylaxis. 6. Using the described methods, smokers can be screened for pre-malignant changes that are indicative of a high risk of developing lung cancer, avoiding repeated exposure to radiation associated with imaging techniques such as CT scans for screening. Epithelial cells can be obtained from the airway epithelium using BAL or other means, combined TGF-β / PI3K-FOXO pathway activity can be performed, and the results can be interpreted in a computer model to provide an airway epithelial abnormality score, indicating the probability of the presence of pre-malignant changes indicative of a risk of developing (squamous) lung cancer. 7. If a smoker presents with symptoms such as a persistent cough, epithelial cells can be obtained from the airway epithelium using BAL or tracheal scraping techniques and combined determination of TGF-β / HH / PI3K-FOXO pathway activity can be performed on RNA isolated from this cell sample and the results can be interpreted in a computer model to provide an airway epithelial abnormality score indicating the probability of the presence of pre-malignant changes indicative of the risk of developing (squamous) lung cancer. 8. If a pulmonary lesion is detected on imaging such as chest x-ray or CT and it is unclear whether this is a malignant lesion, the described tests on airway epithelial cells obtained by BAL / scraping can provide complementary information regarding the nature of the lesion with regard to malignant potential, especially in smokers.
[0136] Application of CDS Referring to FIG. 3 (which diagrammatically illustrates a clinical decision support (CDS) system configured to determine a risk score indicative of a subject's risk of developing airway cancer as disclosed herein), the clinical decision support (CDS) system 10 is implemented as a suitably configured computer 12. The computer 12 is configured to operate as the CDS system 10 by executing suitable software, firmware, or other instructions stored on a hard drive or other magnetic storage medium, an optical disk or other optical storage medium, a random access memory (RAM), a read-only memory (ROM), a flash memory, or other electronic storage medium, a network server, or other non-transitory storage medium (not shown). Although the illustrative CDS system 10 is embodied by an illustrative computer 12, more generally, a CDS system is embodied by a digital processing device or apparatus that includes a digital processor configured to perform the clinical decision support methods described herein. For example, the digital processing device can be a handheld device (e.g., a personal data assistant or a smartphone operating a CDS application), a notebook computer, a desktop computer, a tablet computer or device, a remote network server, or the like. The computer 12 or other digital processing device typically includes or is operably coupled to a display device 14 through which information, including clinical decision aid recommendations, is displayed to the healthcare practitioner. The computer 12 or other digital processing device typically also includes or is operably coupled to an illustrative keyboard 16, or one or more user input devices, such as a mouse, a trackball, a trackpad, a touch-sensitive screen (optionally integrated with the display device 14), or another pointer-based user input device, through which the healthcare practitioner can enter operational commands for controlling the CDS system 10, data to be used by the CDS system 10, or other information.
[0137] The CDS system 10 accepts input information about a subject (e.g., a hospital patient, or an outpatient being treated by an oncologist, clinician, or other medical practitioner, or a person undergoing cancer screening or some other medical diagnosis and known or suspected to have a certain type of airway cancer or a predisposition to developing airway cancer). The CDS system 10 applies various data analysis algorithms to this input information to generate clinical decision aid recommendations that are presented to the medical practitioner via a display device 14 (or via a voice synthesizer or other device that provides a human perceptible output). In some embodiments, these algorithms include applying clinical guidelines to the patient. Clinical guidelines are stored sets of standard or "canonical" treatment recommendations, typically constructed based on the recommendations of a panel of medical professionals, and optionally formatted, in the form of a clinical "flowchart" to facilitate orientation by the clinical guidelines. In various embodiments, the data processing algorithms of the CDS 10 also or instead include various diagnostic or clinical testing algorithms that are performed on the input information to extract clinical decision recommendations, such as the machine learning methods disclosed herein.
[0138] In an illustrative CDS system disclosed herein (e.g., CDS system 10), the CDS data analysis algorithms include one or more diagnostic or clinical testing algorithms that are performed on genomic and / or proteomic input information obtained by one or more medical laboratories 18. These laboratories may be "on-site," i.e., variously located at a hospital or other location where a subject is undergoing medical testing and / or treatment, or may be "off-site," e.g., a specialized or centralized laboratory that receives a subject's sample extracted from the subject (e.g., via mail or another delivery service).
[0139] The sample is processed by a laboratory to generate genomic or proteomic information. For example, the sample is processed using a microarray (variously referred to in the art as gene chip, DNA chip, biochip, etc.) or by quantitative polymerase chain reaction (qPCR) processing that measures telltale genomic or proteomic information such as expression levels of genes of interest, for example in the form of messenger ribonucleic acid (mRNA) levels transcribed from genes or protein levels translated from mRNA transcribed from genes. As another example, the sample is processed by a gene sequencing laboratory to generate deoxyribonucleic acid (DNA) sequences or RNA sequences, copy number variations, methylation, etc. Other contemplated measurement approaches include immunohistochemistry (IHC), cytology, fluorescent in situ hybridization (FISH), proximity ligation assays, etc. performed on pathology slides. Other information that may be generated by microarray processing, mass spectrometry, gene sequencing, or other laboratory techniques includes methylation information. Various combinations of such genomic and / or proteomic measurements may also be performed.
[0140] In some embodiments, the medical laboratory 18 performs some standardized data acquisition on the subject's samples, thereby generating a large amount of genomic and / or proteomic data. For example, the standardized data acquisition techniques generate (optionally aligned) DNA sequences for one or more chromosomes or parts of chromosomes, or for the entire genome. By applying standard microarrays, thousands or tens of thousands of data items can be generated, such as expression levels of a large number of genes, various methylation data, etc. Similarly, PCR-based measurements can be used to measure the expression levels of selected genes. This large amount of genomic and / or proteomic data, or selected parts thereof, is the input of the CDS system 10 to be processed, thereby developing clinically useful information for formulating clinical decision support recommendations.
[0141] The disclosed CDS systems and related methods relate to processing genomic and / or proteomic data to assess the activity of various cell signaling pathways and determine a risk score indicative of a subject's risk of developing airway cancer. However, it should be understood that the disclosed CDS systems (e.g., CDS system 10) optionally further include a variety of additional capabilities, such as generating clinical decision support recommendations according to stored clinical guidelines based on vital sign monitoring data, patient medical history data, patient demographic data (e.g., gender, age, etc.), patient medical imaging data, and various other patient data. Alternatively, in some embodiments, the capabilities of CDS system 10 may be limited to only performing genomic and / or proteomic data analysis to assess the activity of cell signaling pathways and determine a risk score indicative of whether a subject has abnormal airway epithelium and / or is at risk of developing airway cancer, as disclosed herein.
[0142] With continued reference to the exemplary FIG. 3, the CDS system 10 may be configured to detect and / or diagnose cell signaling pathways in a subject based on, but not limited to, expression levels 20 of one or more target genes of the cell signaling pathways measured in a sample from the subject. t , P p , P n ) to predict the activity of two or more cell signaling pathways 22 selected from the group consisting of:
[0143] Measurement of the mRNA expression level of a gene encoding a control protein of a cell signaling pathway, such as an intermediate protein that is part of a protein cascade that forms the cell signaling pathway, is an indirect measure of the expression level of the control protein, which may or may not correlate strongly with the expression level of the actual control protein (but is less correlated with the overall activity of the cell signaling pathway). The cell signaling pathway directly controls the transcription of the target genes, and therefore the expression level of the mRNA transcribed from the target genes is a direct result of this control activity. Thus, the CDS system 10 infers the activity of two or more cell signaling pathways based on the expression level (mRNA or protein level as a surrogate measure) of one or more target genes of the cell signaling pathway. This ensures that the CDS system 10 infers the activity of the pathway based on direct information provided by the measured expression levels of the target genes.
[0144] The predicted activities are then used to determine a risk score, step 24, indicative of the risk of the subject developing respiratory tract cancer, as described in detail herein. The risk score is based on a combination of the predicted activities. For example, the risk score is a "multiple pathway score" (MPS), calculated as described in detail herein and in WO2014174003, WO2016062892, and WO2016062893, each of which is incorporated herein by reference in its entirety for the purposes of calculating the risk score, multiple pathway score (MPS), respectively.
[0145] Based on the determined MPS, the CDS system 10, in this embodiment, assigns the subject to at least one of a plurality of risk groups with various indicated risks for the subject to develop airway cancer (26) and / or determines a recommended treatment for the subject based on the indicated risk (28).
[0146] It is further possible for the CDS system 10 to be configured to combine the risk score with one or more additional risk scores obtained from one or more additional predictive tests to obtain a combination of risk scores, the combination of risk scores being indicative of a subject's risk of developing airway cancer.
Claims
1. 16. A computer-implemented method in which a step of determining whether a subject has abnormal airway epithelium is performed by a digital processing device, said determining step comprising: determining an airway abnormality score indicative of whether the subject has abnormal airway epithelium based on a combination of cell signaling pathway activity in an epithelial cell sample from the subject's airway; The cell signaling pathway comprises two or more cell signaling pathways selected from the group consisting of the TGF-β pathway, the PI3K-FOXO pathway, and the Notch pathway; determining said airway abnormality score further based on a reference activity of each cell signaling pathway of said two or more cell signaling pathways; the reference activities reflect the activity of each cell signaling pathway found in the airway epithelium of a healthy subject; determining the airway abnormality score, deducing quantitative measures of cell signaling pathway activity in said epithelial cell samples obtained from the upper or lower respiratory tract from measuring mRNA levels of three or more confirmed direct target genes of transcription factors associated with said respective cell signaling pathways; determining a cell signaling pathway abnormality score for each of the two or more cell signaling pathways based on the estimated quantitative measurements, and calculating the airway abnormality score by combining the determined cell signaling pathway abnormality scores.
2. 16. A computer-implemented method, wherein determining a risk score indicative of a risk of a subject having abnormal airway epithelium to develop airway cancer is performed by a digital processing device, the determining step comprising: determining said risk score based on a combination of activity of cell signaling pathways in an epithelial cell sample from the airway of said subject; The cell signaling pathway comprises two or more cell signaling pathways selected from the group consisting of the TGF-β pathway, the PI3K-FOXO pathway, and the Notch pathway; determining said risk score further based on a combination of reference activities of said cell signaling pathways; the risk score is defined such that the risk represented by the risk score increases with decreased activity of the TGF-β pathway and increased activity of the PI3K pathway and / or decreased activity of the Notch pathway relative to the reference activity of the cell signaling pathway; determining the risk score, determining an airway abnormality score based on the combination of activity of the cell signaling pathways in the epithelial cell sample and converting the airway abnormality score to the risk score; determining the airway abnormality score, deducing quantitative measures of cell signaling pathway activity in said epithelial cell samples obtained from the upper or lower respiratory tract from measuring mRNA levels of three or more confirmed direct target genes of transcription factors associated with said respective cell signaling pathways; determining a cell signaling pathway abnormality score for each of the two or more cell signaling pathways based on the estimated quantitative measurements, and calculating the airway abnormality score by combining the determined cell signaling pathway abnormality scores.
3. 3. The computer-implemented method of claim 1 or 2, wherein the airway abnormality score and risk score are determined based on evaluation of a mathematical pathway model calibrated to relate activity of the cell signaling pathway in the epithelial cell sample to the airway abnormality score and the risk score, respectively.
4. The computer-implemented method of claim 3, wherein the mathematical pathway model is a Bayesian network model based on conditional probabilities related to activity levels of transcription factor (TF) elements associated with the cell signaling pathway and expression levels of the three or more target genes, or is based on one or more linear combinations of the expression levels of the three or more target genes.
5. The activity of the cell signaling pathway in the epithelial cell sample is receiving expression levels of target genes of each of said cell signaling pathways; determining the activity levels of TF elements associated with cell signaling pathways that control the transcription of said target genes, based on evaluation of a mathematical pathway model calibrated to relate the expression levels of said target genes to the activity levels of said respective cell signaling pathways; inferring an activity of said respective cell signaling pathway based on said determined activity level of a TF element associated with said cell signaling pathway. The computer-implemented method of claim 1 , wherein the determination is made by, or is deducible by, a method comprising:
6. 6. The computer-implemented method of claim 1, wherein the cell signaling pathway comprises the TGF-β pathway and one or both of the PI3K-FOXO pathway and the Notch pathway.
7. the three or more TGF-β target genes are selected from the group consisting of ANGPTL4, CDC42EP3, CDKN1A, CTGF, GADD45A, GADD45B, HMGA2, ID1, IL11, JUNB, PDGFB, PTHLH, SERPINE1, SGK1, SKIL, SMAD4, SMAD5, SMAD6, SMAD7, SNAI2, VEGFA, or the three or more TGF-β target genes are selected from the group consisting of CDC42EP3, GADD45B, HMGA2, ID1, JUNB, OVAL1, VEGFA, SGK1, and / or the three or more PI3K-FOXO target genes are selected from the group consisting of AGRP, BCL2L11, BCL6, BNIP3, BTG1, CAT, CAV1, CCND1, CCND2, CCNG2, CDKN1A, CDKN1B, ESR1, FASLG, FBXO32, GADD45A, INSR, MXI1, NOS3, PCK1, POMC, PPARGC1A, PRDX3, RBL2, SOD2, TNFSF10, and / or The three or more Notch target genes are CD28, CD44, DLGAP5, DTX1, EPHB3, FABP7, GFAP, GIMAP5, HES1, HES4, HES5, HES7, HEY1, HEY2, HEYL, KLF5, MYC, NFKB2, NOX1, NRARP, PBX1, PIN1, PLXND1, PTCRA, SO X9, and TNC, and one or more Notch target genes are selected from the group consisting of CD28, CD44, DLGAP5, EPHB3, FABP7, GFAP, GIMAP5, HES7, HEY1, HEYL, KLF5, NFKB2, NOX1, PBX1, PIN1, PLXND1, SOX9, and TNC; A computer-implemented method according to any one of claims 1 to 6.
8. Based on the determined airway abnormality score, the digital processing device: providing further evidence that the non-diagnostic nodule is malignant or benign; and / or Predicting whether the airway epithelium is pre-malignant; and / or Predicting whether a person is at high risk of developing respiratory tract cancer; and / or Predicting whether a person is at high risk of developing lung cancer; and / or Predicting whether a person is at high risk of developing squamous cell lung cancer; and / or Predicting whether a person is at high risk of developing lung adenocarcinoma; and / or predicting whether a person would benefit from localized treatment to prevent the development of cancer; and / or Predicting whether a patient has lung cancer; and / or Prediction of drug efficacy of chemotherapy and / or hormonal therapy; and / or Monitoring drug efficacy; and / or determining the frequency of monitoring or, more particularly, the frequency of monitoring of the therapeutic response; and / or Drug development, and / or Assay development, and / or Predicting whether a person is at risk of developing invasive airway cancer; and / or predicting whether a person is at risk of disease progression; and / or Predicting or diagnosing whether a person has a reduced risk following treatment; and / or Complementing diagnostic information obtained from imaging and / or other pathological and / or genetic tests; and / or Cancer Staging 2. The computer-implemented method of claim 1, further comprising:
9. 9. An apparatus for determining an airway abnormality score indicative of whether a subject has abnormal airway epithelium or a risk score indicative of a risk of developing airway cancer, comprising a digital processor configured to carry out the computer-implemented method of any one of claims 1 to 8.
10. A non-transitory storage medium storing instructions executable by a digital processing device implementing the computer-implemented method of any one of claims 1 to 8 for determining an airway abnormality score indicating whether a subject has abnormal airway epithelium or a risk score indicative of a risk of developing airway cancer for a subject having abnormal airway epithelium.
11. 9. A computer program for determining an airway abnormality score indicative of whether a subject has abnormal airway epithelium or a risk score indicative of the risk of developing airway cancer, comprising program code means for causing a digital processing device to perform the computer-implemented method of any one of claims 1 to 8 when the computer program is executed on the digital processing device.
12. 1. A kit for determining an airway abnormality score indicative of whether a subject has abnormal airway epithelium or whether a subject with abnormal airway epithelium has a risk score indicative of a risk of developing airway cancer, comprising: Components for determining the expression levels of at least three target genes of the TGF-β cell signaling pathway, at least three target genes of the PI3K-FOXO cell signaling pathway, and at least three target genes of the Notch cell signaling pathway; and A kit comprising the device of claim 9, the non-transitory storage medium of claim 10, or the computer program of claim 11.
Citation Information
Patent Citations
Evaluation of PI3K intracellular signaling pathway activity using mathematical modeling of target gene expression
JP2016536976A
Evaluation of tgf-β cell signaling pathway activity using mathematical modeling of target gene expression
JP2018503354A
Assessment of notch cellular signaling pathway activity using mathematical modelling of target gene expression
WO2019068585A1
Determining functional status of immune cells types and immune response
WO2019068623A1
