Methods for detecting and treating lung cancer

Abstract

Methods for improving the lead time for a patient prior to a diagnosis of lung cancer are provided.

Description

[0001] This application claims priority to U.S. Provisional Application No. 63 / 584,680, filed September 22, 2023, the contents of which are incorporated herein by reference as if written in their entirety.

[0002] This invention was completed with government support under licenses CA200468, CA194733, CA213285, and CA086368 granted by the National Institutes of Health (NIH). The government holds certain rights to this invention.

[0003] Lung cancer is the most prevalent cancer in the United States, with a five-year survival rate of less than 15%. Recently, lung cancer treatment has begun to shift from limited options such as radiation, folate metabolism, platinum-based drugs, and taxane-based drugs to more targeted therapies that require histological characterization of the tumor and / or the presence or absence of key biomarkers or therapeutic target proteins.

[0004] Data from the National Lung Screening Trial (NLST) in the United States indicates that annual screening of high-risk current and former smokers via low-dose computed tomography (LDCT) of the chest can reduce lung cancer mortality by 20%. In 2021, the U.S. Preventive Services Task Force (USPSTF) expanded the eligibility criteria for LDCT screening, now recommending annual LDCT screening for lung cancer in adults aged 50–80 who have smoked more than 20 pack-years and are currently smoking or have quit smoking within the past 15 years. However, in terms of incidence, CT screening has several associated negative aspects, including overdiagnosis, false positives, overtreatment, and financial costs.

[0005] A substantial body of literature on lung cancer risk prediction suggests the potential benefit of supplementing USPSTF screening criteria with risk-based models when identifying candidates for CT screening. For example, recent estimates suggest that 20% of additional lung cancer deaths could be avoided by using screening criteria based on individual risk assessments. The information required to use risk prediction tools can be readily determined by general practitioners—or perhaps through self-assessment using online risk calculators—making it possible that future lung cancer screening programs will implement such tools when assessing screening eligibility.

[0006] One such tool would be risk-based, individual-level screening criteria that accurately estimate a given subject's risk of lung cancer in the near future (e.g., 1-3 years). Several risk prediction models have been published that rely on demographic data (age, sex, etc.) and risk factor data from questionnaires, such as PLCO. m2012And the Liverpool Lung Project (LLP). Elevated levels of protein biomarkers have also been found to be useful predictors of lung cancer development risk. US 16 / 484,177 describes a novel blood-based four-marker proteome comprising or composed of the following: presurfactant protein B (pre-SFTPB), mucin 16 (CA125), carcinoembryonic antigen (CEA), and cytokeratin-19 fragment (CYFRA21-1), the contents of which are hereby incorporated by reference in their entirety. It has been found that the use of this proteome significantly improves lung cancer risk assessment compared to previous and current USPSFT criteria used for lung cancer screening.

[0007] Therefore, a method or test is needed to improve the lead time before lung cancer diagnosis. Longitudinal screening methods have been found to provide improved performance compared to single measurements. These longitudinal screening methods involve sequential testing of a four-marker proteome comprising presurfactant protein B (pre-SFTPB), mucin 16 (CA125), carcinoembryonic antigen (CEA), and cytokeratin-19 fragment (CYFRA21-1), or a combination thereof, using a parametric empirical Bayesian (PEB) algorithm. Summary of the Invention

[0008] This article provides a method for determining a patient's risk of developing lung cancer, the method comprising: The model score is calculated using two or more biomarker scores collected at two or more time points, where each biomarker score is determined by the levels of biomarkers CEA, CA125, CYFRA21-1, and Pro-SFTPB in biological samples obtained from the patient at each time point; and The patient is identified as being at risk of lung cancer or not by comparing the model score with predefined parameters.

[0009] This article also provides a method for improving the lead time before lung cancer diagnosis, which includes: The model score is calculated using two or more biomarker scores collected at two or more time points, where each biomarker score is determined by the levels of biomarkers CEA, CA125, CYFRA21-1, and Pro-SFTPB in biological samples obtained from the patient at each time point; and The patient is identified as being at risk of lung cancer or not by comparing the model score with predefined parameters.

[0010] This article also provides a method that includes: The model score is calculated using two or more biomarker scores collected at two or more time points, wherein each biomarker score is determined by the level of biomarkers CEA, CA125, CYFRA21-1 and Pro-SFTPB in biological samples obtained from asymptomatic lung cancer patients at each time point. By comparing the model score with predefined parameters, the patient was identified as being at risk for lung cancer; and The patient, who was identified as being at risk for lung cancer, underwent a computed tomography (CT) scan.

[0011] This article also provides a method for identifying and treating asymptomatic patients at increased risk of lung cancer, the method comprising: The model score is calculated using two or more biomarker scores collected at two or more time points, wherein each biomarker score is determined by the level of biomarkers CEA, CA125, CYFRA21-1 and Pro-SFTPB in biological samples obtained from the patient at each time point; By comparing the model score with predefined parameters, the patient was identified as being at risk for lung cancer. The patient, identified as being at risk for lung cancer, underwent a computed tomography (CT) scan; and The cancerous tumor identified in the CT scan was surgically removed. Attached Figure Description

[0012] Figure 1 ROC curves were plotted to evaluate the 4MP performance using the Parametric Empirical Bayes (PEB) and Single Threshold (ST) methods.

[0013] Figure 2 ROC curves were plotted to evaluate the performance of 4MP using parametric empirical Bayes (PEB) and single threshold (ST) methods in early lung cancer.

[0014] Figure 3 ROC curves were plotted to evaluate the 4MP performance using parametric empirical Bayes (PEB) and single threshold (ST) methods in advanced lung cancer.

[0015] Figure 4 ROC curves were plotted to evaluate the 4MP performance using parametric empirical Bayes (PEB) and single threshold (ST) methods in high-risk smoking layers.

[0016] Figure 5 ROC curves were plotted to evaluate the 4MP performance using the Parametric Empirical Bayes (PEB) and Single Threshold (ST) methods in a medium-risk smoking layer.

[0017] Figure 6ROC curves were plotted to evaluate the 4MP performance using parametric empirical Bayes (PEB) and single threshold (ST) methods in a low-risk smoking layer.

[0018] Figure 7 ROC curves were plotted to evaluate the 4MP performance using parametric empirical Bayes (PEB) and single threshold (ST) methods in adenocarcinoma and lung cancer.

[0019] Figure 8 ROC curves were plotted to evaluate the 4MP performance in squamous lung cancer using parametric empirical Bayes (PEB) and single threshold (ST) methods.

[0020] Figure 9 ROC curves were plotted to evaluate the 4MP performance in small cell lung cancer using parametric empirical Bayes (PEB) and single threshold (ST) methods.

[0021] Figure 10 The study depicted PEB and ST positivity at a 1.7% 6-year specificity threshold, where both methods produced positive results on the first biomarker measurement tested. N=222.

[0022] Figure 11 PEB and ST positivity were depicted at a 6-year specificity threshold of 1.7%, where neither method produced a positive test result. N=27.

[0023] Figure 12 The study depicted PEB and ST positivity at a 1.7% 6-year specificity threshold, where the first positive result did not come from the first biomarker measurement, or PEB or ST was positive, but not both. N=75.

[0024] Figure 13 PEB and ST positivity at a 1.0% 6-year specificity threshold were depicted, with both methods producing positive results on the first biomarker measurement tested. N=261.

[0025] Figure 14 The duration (in years) of different positive signals for the PEB and ST methods at a 1.0% 6-year specificity threshold is depicted. A) Histogram distribution of the two methods. B) Individual time representation of positive results. "+" indicates no detection by the PEB / ST method.

[0026] Figure 15 PEB and ST positivity at a 1.0% 6-year specificity threshold were depicted, where the first positive result did not come from the first biomarker measurement, or PEB or ST was positive, but not both. N=51. Detailed Implementation

[0027] This article provides a method for determining a patient's risk of developing lung cancer, the method comprising: The model score is calculated using two or more biomarker scores collected at two or more time points, where each biomarker score is determined by the levels of biomarkers CEA, CA125, CYFRA21-1, and Pro-SFTPB in biological samples obtained from the patient at each time point; and The patient is identified as being at risk of lung cancer or not by comparing the model score with predefined parameters.

[0028] This article also provides a method for improving the lead time before lung cancer diagnosis, which includes: The model score is calculated using two or more biomarker scores collected at two or more time points, where each biomarker score is determined by the levels of biomarkers CEA, CA125, CYFRA21-1, and Pro-SFTPB in biological samples obtained from the patient at each time point; and The patient is identified as being at risk of lung cancer or not by comparing the model score with predefined parameters.

[0029] This article also provides a method that includes: The model score is calculated using two or more biomarker scores collected at two or more time points, wherein each biomarker score is determined by the level of biomarkers CEA, CA125, CYFRA21-1 and Pro-SFTPB in biological samples obtained from asymptomatic lung cancer patients at each time point. By comparing the model score with predefined parameters, the patient was identified as being at risk for lung cancer; and The patient, who was identified as being at risk for lung cancer, underwent a computed tomography (CT) scan.

[0030] This article also provides a method for identifying and treating asymptomatic patients at increased risk of lung cancer, the method comprising: The model score is calculated using two or more biomarker scores collected at two or more time points, wherein each biomarker score is determined by the level of biomarkers CEA, CA125, CYFRA21-1 and Pro-SFTPB in biological samples obtained from the patient at each time point; By comparing the model score with predefined parameters, the patient was identified as being at risk for lung cancer. The patient, identified as being at risk for lung cancer, underwent a computed tomography (CT) scan; and The cancerous tumor identified in the CT scan was surgically removed.

[0031] In some embodiments, the model score is calculated using the following equation: in: Y is the biomarker score at the most recent time point, and is expressed as 0.4730. log[CA125] + 0.6531 log[CEA] + 0.2612 log[CYFRA21-1] + 0.9238 Calculate log[Pro-SFTPB]; µ is approximately 7.07; V is approximately 0.2672; B1 is approximately 0.767; B n Using equation (n) B1) / (n The calculation is B1 + (1 – B1)), where n is the total number of tests performed on the patient; and It is the average Y score calculated from biomarkers collected up to the most recent point in time.

[0032] In some embodiments, the predefined parameter is approximately 0.63.

[0033] In some embodiments, the predefined parameter is approximately 0.45.

[0034] In some embodiments, a model score greater than a predefined parameter is considered a positive test.

[0035] In some embodiments, a model score less than a predefined parameter is considered a negative test.

[0036] In some embodiments, biomarker scores are collected at two to five time points.

[0037] In some embodiments, biomarker scores are collected at two time points.

[0038] In some embodiments, biomarker scores are collected at three time points.

[0039] In some embodiments, biomarker scores are collected at four time points.

[0040] In some embodiments, biomarker scores are collected at five time points.

[0041] In some embodiments, the individual is asymptomatic.

[0042] In some embodiments, the levels of biomarkers CEA, CA125, CYFRA21-1, and Pro-SFTPB in biological samples obtained from individuals are determined by immunoassay.

[0043] In some embodiments, each of the biomarkers CEA, CA125, CYFRA21-1, and Pro-SFTPB in a biological sample obtained from an individual generates a detectable signal.

[0044] In some embodiments, the detectable signal can be detected by spectroscopy.

[0045] In some embodiments, the spectroscopic method is selected from ultraviolet-visible spectroscopy, mass spectrometry, nuclear magnetic resonance (NMR) spectroscopy, proton NMR spectroscopy, gas chromatography-mass spectrometry (GC-MS), liquid chromatography-mass spectrometry (LC-MS), correlation spectroscopy (COSY), nuclear Overhouse effect spectroscopy (NOESY), nuclear Overhouse effect spectroscopy in rotating coordinates (ROESY), time-of-flight LC-MS (LC-TOF-MS), liquid chromatography-tandem mass spectrometry (LC-MS / MS), and capillary electrophoresis-mass spectrometry.

[0046] In some embodiments, the spectroscopy method is mass spectrometry.

[0047] In some embodiments, the mass spectrometry is LC-TOF-MS.

[0048] In some embodiments, the lung cancer is early-stage (e.g., stage I or stage II).

[0049] In some embodiments, the lung cancer is advanced (e.g., stage III or IV).

[0050] In some embodiments, the individual has a smoking history of ≥10 pack-years.

[0051] In some embodiments, the individual is between 50 and 80 years old.

[0052] In some embodiments, the method has an AUC greater than that of different single biomarkers, multiple biomarkers, groups, assays, algorithms, models, or any combination thereof.

[0053] In some embodiments, the AUC is greater than 0.81.

[0054] In some embodiments, the AUC is between 0.83 and 0.89.

[0055] In some embodiments, the AUC is approximately 0.86.

[0056] In some embodiments, the method has greater sensitivity and specificity at a risk threshold of ≥1.7% / 6 years than different single biomarkers, multiple biomarkers, groups, assays, algorithms, models, or any combination thereof.

[0057] In some embodiments, the sensitivity is greater than 0.81 and the specificity is about 0.63.

[0058] In some embodiments, the sensitivity is between 0.84 and 0.98.

[0059] In some embodiments, the sensitivity is approximately 0.91.

[0060] In some embodiments, the method has greater sensitivity and specificity at a risk threshold of ≥1.0% / 6 years than different single biomarkers, multiple biomarkers, groups, assays, algorithms, models, or any combination thereof.

[0061] In some embodiments, the sensitivity is greater than 0.91 and the specificity is about 0.45.

[0062] In some embodiments, the sensitivity is between 0.94 and 0.99.

[0063] In some embodiments, the sensitivity is approximately 0.96.

[0064] In some embodiments, the lead time before diagnosis of lung cancer patients is greater than the lead time before diagnosis for different individual biomarkers, multiple biomarkers, groups, assays, algorithms, models, or any combination thereof.

[0065] In some embodiments, the lead time prior to diagnosis is greater than 1.03 years.

[0066] In some embodiments, the lead time prior to diagnosis is between 1.26 and 2.70 years.

[0067] In some embodiments, the patient is subsequently given further lung cancer screening or treatment.

[0068] In some embodiments, screening is selected from endoscopic ultrasound, magnetic resonance imaging (MRI), and computed tomography (CT).

[0069] In some embodiments, screening is performed annually.

[0070] In some embodiments, screening is performed every six months.

[0071] In some embodiments, the treatment is selected from surgery, chemotherapy, immunotherapy, radiotherapy, targeted therapy, or combinations thereof. definition

[0072] As used herein, the following terms have the meanings indicated.

[0073] When disclosing ranges of values ​​and using the symbols “n1 … to n2” or “between n1 … and n2” (where n1 and n2 are numbers), unless otherwise stated, this symbol is intended to include the numbers themselves and the range between them. This range can be an integer range between and including the endpoints, or a continuous range. For example, the range “2 to 6 carbons” is intended to include two, three, four, five, and six carbons, since carbon appears in integer units. For example, comparing the range “1 to 3 µM (micromoles)” (which is intended to include 1 µM, 3 µM, and all numbers in between) to any number with significant figures (e.g., 1.255 µM, 2.1 µM, 2.9999 µM, etc.) is also meaningful.

[0074] As used herein, the term “about” is intended to define the numerical value it modifies, indicating that such a value is variable within a range. Unless a specific range is listed (such as an error limit or the standard deviation of the mean given in a chart or table), the term “about” should be understood to mean the larger of the range covering the listed value, the range including significant figures by rounding to that number, and the range covering ±20% of the listed value.

[0075] As used herein, “lung cancer” refers to a malignant growth in the lung characterized by abnormal cell proliferation, in which cell growth exceeds and is out of harmony with the growth of the surrounding normal tissue. In some embodiments, the severity of lung cancer may vary, indicated as stages I through IV. In some embodiments, lung cancer may be in an early stage (e.g., stage I or II) or it may be in an advanced stage (e.g., stage III or IV).

[0076] As used herein, the terms “subject” or “patient” refer to a mammal, preferably a human, that needs to be classified as lung cancer positive or lung cancer negative and that can be provided with further treatment.

[0077] As used herein, “healthy” refers to an individual for whom no evidence of lung cancer has been found, i.e., the individual does not have lung cancer. Such individuals may be classified as “lung cancer negative” or as having healthy lungs or normal, unimpaired lung function. Healthy patients or subjects do not have symptoms of lung cancer but may have benign lung nodules or masses, i.e., a combination of adenomas and cysts, or one or more non-cancerous lung conditions, such as chronic obstructive pulmonary disease (COPD). In some embodiments, healthy patients or subjects may be used for comparison with diseased or suspected diseased samples to identify lung cancer in patients or patient groups.

[0078] As used herein, the terms "treating" or "treatment" refer to the application of a therapeutic method to an individual who has presented or previously presented at least one symptom of a disease or condition. For example, "treatment" can include reducing, weakening, or improving symptoms of a disease or condition; preventing additional symptoms; improving the underlying metabolic cause of symptoms; inhibiting a disease or condition, for example, preventing its progression; alleviating a disease or condition; causing its remission; alleviating symptoms caused by the disease or condition; or stopping the symptoms of a disease or condition. For example, the term "treatment" in relation to a disorder means reducing the severity of one or more symptoms associated with that particular disorder. Therefore, treating a disorder does not necessarily mean reducing the severity of all symptoms associated with the disorder, nor does it necessarily mean completely reducing the severity of one or more symptoms associated with the disorder. In connection with this disclosure, the term can also mean the administration of a pharmacological substance or preparation, or the performance of non-pharmacological methods, including but not limited to radiation therapy and surgery. Pharmacological substances as used herein may include, but are not limited to, anticancer drugs, including chemotherapy drugs, polyamine inhibitors, hormone therapy, and targeted therapy. Examples of chemotherapy drugs for lung cancer include paclitaxel / taxool (e.g., albumin-bound paclitaxel or albumin-conjugated paclitaxel, brand name Abraxane®), erlotinib (Tarceva®, etc.), afatinib (Gilotrif®), gefitinib (Iressa®), bevacizumab (Avastin®), gemcitabine (Gemzar®), crizotinib (Xalkori®), ceritinib (Zykadia®), cisplatin / cisplatin, carboplatin (Paraplatin®), docetaxel (Taxotere®), pemetrexed (Alimta®), and vinorelbine (Navelbine®); as well as combination chemotherapy regimens, including cisplatin + Paclitaxel, TIP (paclitaxel / paclitaxel, ifosfamide, and cisplatin / cisplatin), VeIP (vincrine, ifosfamide, and cisplatin / cisplatin), VIP (etoposide / VP-16, ifosfamide, and cisplatin / cisplatin), VAC (vincristine, cytosine, and cyclophosphamide), and PEB (cisplatin / cisplatin, etoposide, and bleomycin). The terms "pharmacological substance" and "anticancer therapy" may also include substances used in immunotherapy, such as checkpoint inhibitors. Treatment may include a variety of pharmacological substances or a variety of treatment methods, including but not limited to surgery and chemotherapy.

[0079] As used herein, “amount” or “level” refers to a generally quantifiable measurement of the biomarker described herein, wherein the measurement enables comparison of the biomarker between samples and / or with a control sample. In some embodiments, a amount or level is quantifiable and refers to the level of a specific biomarker in a biological sample (e.g., blood, serum, urine, etc.), as determined by laboratory methods or tests such as immunoassays (e.g., antibodies), mass spectrometry, or liquid chromatography. In some embodiments, the biomarker may be present in the sample in an increased or decreased amount. Biomarker comparisons may be based on direct measurements of biomarker levels as described herein (e.g., by protein quantification or gene expression analysis) or on measurements such as reporter molecules, biomarker-receptor complexes, biomarker-relay-receptor complexes, etc.

[0080] As used herein, the term "elevated" refers to a biomarker level or model score in a given subject that is higher than the same biomarker level or model score in a given group of healthy patients or subjects. In some embodiments, elevated PLCO m2012 The model score is 0.00948 or higher. In some embodiments, the PLCO is increased. m2012 The model score is 0.016082 or higher.

[0081] As used herein, the term "hazard ratio" refers to a measure of how frequently a particular event occurs in one group compared to how frequently it occurs in another group over time. Hazard ratios are commonly used in clinical trials to measure the survival rate at any given time point between a group of patients receiving a specific treatment and a control group receiving another treatment or a placebo. The hazard ratio is defined as the slope of the survival curve—a measure of the rate of death of subjects. A hazard ratio of 1 means there is no difference in survival rates between the two groups. A hazard ratio greater than 1 or less than 1 means that one group has a better survival rate. If the hazard ratio is 2.0, the mortality rate in one treatment group is twice that of the other.

[0082] As used herein, the term "regression" refers to a statistical method that assigns predicted values ​​to latent characteristics of a sample based on observable traits (or a set of observable traits). In some embodiments, the characteristics are not directly observable. For example, the regression method used herein can correlate the qualitative or quantitative results of a particular biomarker test or set of biomarkers for a subject with the probability that the subject is positive for lung cancer.

[0083] As used herein, the term "biomarker score" refers to the numerical score of a given biomarker or set of biomarkers measured in a subject's sample. Biomarker scores are calculated by normalizing or weighting the measured levels using fixed coefficients prescribed by a statistical method for a given set of biomarkers. Biomarker scores are used as a component in calculating a subject's model score.

[0084] As used herein, “use” of biomarkers for the diagnosis of lung cancer means the quantification of the level or amount of one or more biomarkers described herein in a biological sample. Quantification can be performed using any method or technique known in the art or described herein. In some embodiments, biomarkers may be used or combined as a group for statistical comparison with other samples.

[0085] As used herein, the term "classification" refers to assigning a subject to either a risk level or a risk level of lung cancer based on the results of biomarker scores, risk scores, or risk profiles obtained for that subject.

[0086] As used herein, the term "sensitivity," in the context of various biochemical assays, refers to the ability to correctly identify individuals with disease (i.e., the true positive rate). By comparison, the term "specificity," as used herein, refers to the ability to correctly identify individuals without disease (i.e., the true negative rate) in the context of various biochemical assays. Sensitivity and specificity are statistical measures of the performance of binary classification tests (i.e., classification functions). Sensitivity quantifies the avoidance of false negatives, while specificity has the same effect on false positives.

[0087] As used herein, “sample” means a test substance in which the presence, level, or concentration of a biomarker as described herein is to be tested. A sample may be any suitable substance according to this disclosure, including but not limited to blood, serum, plasma, or any portion thereof.

[0088] As used in this paper, the term "ROC" refers to Receiver Operating Characteristic, which is a graphical representation used in this paper to measure the performance of a diagnostic method at various cutoff points. The ROC plot can be constructed from the scores of true positives and false positives at different cutoff points.

[0089] As used in this article, the term "AUC" refers to the area under the ROC curve. AUC can be used to estimate the predictive power of a diagnostic test. Generally, the larger the AUC, the stronger the predictive power and the lower the frequency of prediction errors. The possible values ​​of AUC range from 0.5 to 1.0, where a value of 1.0 represents an error-free prediction method.

[0090] As used in this article, the term "p-value" or "p" refers to the probability that, in the context of the Wilcoxon rank-sum test, the biomarker scores of lung cancer-positive and lung cancer-negative subjects are distributed in the same way. Generally, a p-value close to zero indicates that a particular statistical method has a high predictive power when classifying subjects.

[0091] As used herein, the term "CI" refers to a confidence interval, that is, an interval in which a value can be predicted at a certain confidence level. As used herein, the term "95% CI" refers to an interval in which a value can be predicted at a 95% confidence level.

[0092] As used in this article, the term "positive predictive value" refers to the proportion of true positives among positive results obtained by a certain method.

[0093] As used in this article, the term “disease progression” or “early disease progression” is defined as an upgrade in Gleason score and / or an increase in tumor volume at the time of monitoring biopsy within 18 months of the start of active surveillance.

[0094] The phrase “therapeuticly effective” is intended to limit the amount of active ingredient used in the treatment of a disease or disorder or in relation to clinical endpoints. List of abbreviations

[0095] 4MP = Four marker proteome (presurfactant B (preSFTPB), mucin 16 (CA125), carcinoembryonic antigen (CEA), and cytokeratin-19 fragment (CYFRA21-1)); AUC = Area under the curve; ROC = Recipient operating characteristics. Example

[0096] The following examples illustrate embodiments of this disclosure. These examples are presented by way of illustration only and to assist those skilled in the art in using this disclosure. These examples are not intended to limit the scope of this disclosure in any way. Those skilled in the art will understand from this disclosure that many changes can be made to the specific embodiments disclosed and still obtain the same or similar results without departing from the spirit and scope of the invention. Example 1: PLCO Sample Set PLCO queue

[0097] The PLCO Cancer Screening Trial is a randomized, multicenter trial in the United States designed to evaluate the impact of early detection programs for prostate, lung, colorectal, and ovarian cancers on disease-specific mortality. A biobank was created annually from blood samples collected from consenting intervention group participants. Cancer status was reported based on an annual questionnaire. Medical records were obtained to document diagnostic follow-up and characteristics of any diagnosed lung cancer. TNM staging and staging groups were determined using the American Joint Committee on Cancer Staging Manual, Fifth Edition. Treatment data were extracted from medical records one year after diagnosis. After the PLCO study concluded, PLCO participants were followed up for 13 years for lung cancer incidence and 20 years for lung cancer mortality.

[0098] All deaths occurring during the trial were primarily identified through the annual study update questionnaire. Participants who did not return the questionnaire were contacted via repeat mail or telephone. To enhance the integrity of endpoint validation, active follow-up was accompanied by regular correlation with the National Mortality Index. Death certificates were obtained to confirm death and determine the provisional cause of death. Because the underlying cause of death is not always accurately recorded on death certificates, the PLCO trial used an endpoint adjudication procedure to designate the cause of death in a uniform and impartial manner. All deaths whose causes were likely related to cancer were reviewed by a death review board consisting of a non-voting chairperson and three experienced reviewers. The death reviewers were unaware of the deceased participant's trial group. Lung cancer-specific deaths were defined as those whose underlying cause was lung cancer or those resulting from lung cancer treatment.

[0099] All histologically confirmed lung cancers from former smokers with at least two blood draws in the intervention group were included in this study. These participants were diagnosed within six years of entry into the study and had at least one biomarker measurement within two years of diagnosis (n=338 case participants). Non-case participants who had been former smokers and had at least two blood draws were randomly selected (n=2,432 non-case participants). (Table 1). Participants with ≥10 PY of smoking experience were selected as the intended screening population (Table 2). Table 1. Number of participants by the number of continuous biomarker measurements Table 2. Number of participants by the number of continuous biomarker measurements (smoking ≥10 PY) Determination of 4MP in the PLCO sample set

[0100] In the absence of a case-control status, samples from all study participants used for training and testing were sent to the MD Anderson Cancer Center laboratory on dry ice, where they were maintained below -80°C until analysis. Concentrations of pre-SFTPB, CA125, CEA, and CYFRA21-1 were determined using a bead-based immunoassay on a MAGPIX® instrument (Luminex Corporation, Austin, Texas). Samples were analyzed in batches of 36, in duplicate, with matched cases and controls in the same batch in randomized order. Quality control procedures included seven calibration standards, two quality control samples, and one blank sample per batch, in duplicate. Intra-batch and inter-batch coefficients of variation (CV) were 6.86% and 15.54% (for CA125), 1.45% and 9.32% (for CEA), 6.55% and 17.26% (for pre-SFTPB), and 5.56% and 28.71% (for CYFRA21-1), respectively. Biomarker scores for 4MP were derived using fixed beta coefficients from a previously developed logistic regression model. See US 16 / 484,177. The coefficients of variation (CV) for pre-SFTPB, CA125, CEA, and CYFRA21-1 in quality control samples were 22.2%, 12.8%, 10.8%, and 22.6%, respectively. Statistical analysis

[0101] A single-threshold (ST) approach is considered, which compares only the current biomarker measurement for each individual in the population to the same threshold, while the Parametric Empirical Bayes (PEB) algorithm adjusts the biomarker threshold at each test to reflect the participant's medical history. In short, the PEB algorithm utilizes a simple model of biomarker values ​​in the non-case population to estimate the overall mean biomarker value (μ) and the variability of measurements within and between non-case participants. These parameters are used to centralize and scale the biomarker measurements, comparing the current measurement to μ and a weighted average of the participant's previous biomarker values. As the number of repeated biomarker measurements increases, the PEB reference level, and therefore the test threshold, becomes increasingly personalized. In contrast, the ST approach compares the current biomarker measurement to a non-case reference level μ. In both approaches, large deviations from the corresponding reference levels indicate that these values ​​are abnormally high in the non-case population.

[0102] For this study, the false positive rate (FPR) was estimated at the screening level, defined as the proportion of positive results among all screenings performed in the control group. Screening level specificity was defined as 1-FPR. The true positive rate (TPR) or sensitivity (defined as the proportion of lung cancer cases “tested” with at least one positive biomarker) at the patient level was estimated in accordance with previous applications of this method. A threshold was selected to span the full range of false positive rates and to plot ROC curves of TPR versus FPR, and the area under the curve (AUC) was calculated.

[0103] PEB parameters were estimated from non-case participants with a smoking history of ≥10 PY (Table 3). Table 3. Population parameters estimated from individuals with a 10-pack-year smoking history, who have undergone at least two biomarker measurements, and who have not been diagnosed with lung cancer.

[0104] To account for outcome-dependent sampling and multiple measurements for each participant, stratified and cluster bootstrapping methods were used to calculate the 95% CI as percentiles of the sampling distribution estimated from 1,000 resamples. Analysis was performed using R software version 4.2.0 (R Project for Statistical Computation). The population mean (μ), variance (V), and within-class correlation (ICC or B) were estimated using a linear random effects model. Specifically, X ij Equals the 4MP biomarker score of person i at screening times j=1...n and follows a statistical model. The average value per person is µ i Changes within a group The model shows that a single (n=1) measurement has a population mean. variance: and ICC

[0105] ICC measures the similarity of biomarker levels between individuals in an intra-individual comparison. A multilevel package in R software is used to estimate B1. At each screening, a PEB bias model score is calculated for each patient by comparing the current 4MP biomarker score (Y) to a function that includes the sample mean of the patient's n≥0 previous 4MP scores, denoted as . The ICC (B) of the average value of this sample. n ) calculated as

[0106] The PEB model score is calculated as follows:

[0107] A "test" is considered positive when the model score is greater than a predetermined cutoff parameter, where the cutoff value can be estimated empirically based on the training data or by using the percentiles of a standard normal distribution. Predictive performance of the PEB model for diagnosing lung cancer

[0108] In individuals ≥10 PY, the PEB algorithm considering 4MP repeated measures improved the area under the curve (AUC) by 5% compared to the ST method. PEB 0.86 relative to AUC ST (0.81; P-value < 0.05) Figure 1 The benefit of the PEB algorithm was observed in cases stratified as clinically diagnosed with early (I+II) or advanced (III-IV) lung cancer, with an AUC improvement of 0.06 (AUC 0.06) in each case. PEB 0.83 relative to AUC ST : 0.77; P value < 0.05) and 0.04 (AUC) PEB 0.89 relative to AUC ST (P < 0.05) Figure 2-9 ). Figure 4-6 A comparison of PEB and ST is provided in the low, medium and high risk stratification.

[0109] At a pre-established specificity threshold of 63.2% (corresponding to the specificity of 4MP at a 1.7% 6-year risk threshold), PEB improved sensitivity by 9.8% (Sen PEB 90.7% relative to Sen ST The p-value was 80.9; P < 0.05, which is equivalent to detecting 48.7% of cases missed by ST (Table 4). Of the 324 cases ≥ 10 PY, 297 (92.6%) were positive “tests” by either PEB or ST. Of these 297, 222 (74.7%) had a positive PEB / ST “test” result in the first biomarker measurement. Figure 10 and Figure 11 Of the 75 cases that initially tested negative at the first blood draw, 45 (60.0%) received an earlier positive test based on the PEB algorithm compared to the ST method, with a mean lead time of 1.21 (IQR: 0.63–2.07). Figure 12 (Table 5). Table 4. Sensitivity of PEB and ST methods under predefined specificity. 1 This corresponds to the USPSTF2021 standard. 2This corresponds to the USPSTF 2013 standard. Table 5. Precedence time of PEB relative to ST in different subgroups (in years) at a 1.7% 6-year risk-specific threshold (63.2%).

[0110] At a specificity threshold of 45.4%, corresponding to a 1% 6-year risk (corresponding to the current USPSTF 2021 screening guidelines), PEB improves sensitivity by 5.3% (Sen PEB 96.3% relative to Sen ST : 91.0; P value < 0.05 (Table 4). Of the 324 cases ≥ 10 PY, 312 (96.3%) were positive “tests” by PEB or ST. Of these 312, 261 (83.6%) had a positive PEB / ST “test” result in the first biomarker measurement ( Figure 13 Of the 51 cases that initially tested negative at the first blood draw, 27 (52.9%) received an earlier positive test based on the PEB algorithm compared to the ST method, with a mean lead time of 1.37 (IQR: 0.97–1.83). Figure 14 and 15 (Table 6). Table 6. Precedence time of PEB relative to ST in different subgroups (in years) at a 1.0% 6-year risk-specific threshold (45.4%).

[0111] At a pre-established specificity threshold of 45.4% (corresponding to a 1.0% 6-year lung cancer risk), among the 28 individuals with positive PEB or ST signals, the PEB algorithm showed positive results in 17 individuals, with a lead time of 1.26 years prior to diagnosis (IQR: 0.87–2.15), while ST remained negative (Table 7). According to USPSTF 2021 criteria, 6 of the 17 individuals (35.3%) were ineligible for LDCT screening. Ten individuals with positive ST results and a mean lead time of 1.03 years (IQR: 0.27–1.69) were positive based on PEB results, with a mean lead time of 2.70 years (IQR: 2.02–3.54). Figure 14 (Tables 7-9 and 12).

[0112] At a pre-established specificity threshold of 63.2% (corresponding to a 1.7% 6-year lung cancer risk), among 50 individuals with positive PEB or ST signals, a positive result from the PEB algorithm would trigger CT screening in 35 ST-negative individuals, with a mean lead time of 1.19 years prior to diagnosis (interquartile range (IQR): 0.81–1.92). According to the USPSTF 2013 eligibility criteria, 17 of these individuals (48.6%) were ineligible for LDCT screening (Table 7). Ten individuals with positive ST signals and a mean lead time of 0.90 years (IQR: 0.46–1.48) were also eligible for PEB-positive results, with a mean lead time of 3.35 years (IQR: 2.81–3.67). Figure 7 , 10 -12). Table 7. Lead time estimation by PEB and ST methods under predefined specificity in individuals with a smoking history of ≥10 PY. N represents the number of participants in the case; abbreviation: IQR - interquartile range. Table 8. Lead time estimates for PEB and ST methods in different risk strata at a predefined 1.0% 6-year risk specificity threshold. N represents the number of participants in the case; abbreviation: IQR - interquartile range. Table 9. Lead time estimates for PEB and ST methods in case participants stratified by stage and histological subtype at a predefined 1.0% 6-year risk-specific threshold. N represents the number of participants in the case; abbreviation: IQR - interquartile range. Table 10. Lead time estimates for PEB and ST methods in different risk strata at a predefined 1.7% 6-year risk specificity threshold. N represents the number of participants in the case; abbreviation: IQR - interquartile range. Table 11. Lead time estimates for PEB and ST methods in case participants stratified by stage and histological subtype at a predefined 1.7% 6-year risk-specific threshold. N represents the number of participants in the case; abbreviation: IQR - interquartile range. Table 12. Distribution of positive test results by PEB and ST methods in non-case participants.

[0113] Compared to the ST method, continuous measurement of 4MP within the context of an adaptive PEB algorithm model improves the sensitivity and lead time for lung cancer risk assessment in lung cancer screening. Testing for 4MP is useful for individuals currently eligible for LDCT screening and can be extended to those with a smoking history ≥10 PY, enabling risk-based referral to LDCT screening through collaborative decision-making. It is recommended that individuals with an initial "negative" test undergo repeat testing at intervals matched to their risk level. This creates a situation where information from repeated measurements can provide additional information about lung cancer risk.

[0114] The advantage of PEB is its ability to adjust for low baseline 4MP values, thus allowing for the identification of individuals at risk of lung cancer despite ST-seropositivity. For this purpose, volume doubling time (VDT) (defined as the number of days it takes for a nodule volume to double) is a clinically important indicator in lung cancer screening, typically ranging from 20 days to <590 days for malignant nodules. Changes in biomarker levels may reflect the spread from occult tumors into the bloodstream during cancer proliferation and progression. In this instance, at a 1% 6-year risk threshold, the PEB algorithm resulted in 43% of patients subsequently diagnosed with lung cancer receiving a positive “test” earlier, with a mean lead time of 1.37 years, which is expected to lead to benefits in staging and reduced mortality.

[0115] All references, patents, or applications cited in this application, whether U.S. or foreign, are incorporated herein by reference as if they were written in their entirety herein. In case of any discrepancy, the material actually disclosed herein shall prevail.

[0116] Based on the above description, those skilled in the art can easily determine the essential features of the present invention and can make different changes and modifications to the present invention without departing from the spirit and scope of the present invention, so as to adapt it to different uses and conditions.

Claims

1. A method for determining a patient's risk of developing lung cancer, the method comprising: The model score is calculated using two or more biomarker scores collected at two or more time points, wherein each biomarker score is determined by the level of biomarkers CEA, CA125, CYFRA21-1 and Pro-SFTPB in biological samples obtained from the patient at each time point; as well as By comparing the model score with predefined parameters, the patient is identified as being at or not at risk of lung cancer.

2. A method for improving the lead time before diagnosis in lung cancer patients, the method comprising: The model score is calculated using two or more biomarker scores collected at two or more time points, where each biomarker score is determined by the levels of biomarkers CEA, CA125, CYFRA21-1, and Pro-SFTPB in biological samples obtained from the patient at each time point; and The patient was identified as having an increased risk of lung cancer by comparing the model score with predefined parameters.

3. A method comprising: The model score is calculated using two or more biomarker scores collected at two or more time points, wherein each biomarker score is determined by the level of biomarkers CEA, CA125, CYFRA21-1 and Pro-SFTPB in biological samples obtained from the patient at each time point; The patient is identified as being at risk of lung cancer or not by comparing the model score with predefined parameters. as well as The patient, who was identified as being at risk for lung cancer, underwent a computed tomography (CT) scan.

4. A method for identifying and treating asymptomatic patients at increased risk of lung cancer, the method comprising: The model score is calculated using two or more biomarker scores collected at two or more time points, wherein each biomarker score is determined by the level of biomarkers CEA, CA125, CYFRA21-1 and Pro-SFTPB in biological samples obtained from the patient at each time point; The patient is identified as being at risk of lung cancer or not by comparing the model score with predefined parameters. The patient, who was identified as being at risk of lung cancer, underwent a computed tomography (CT) scan. as well as The cancerous tumor identified in the CT scan was surgically removed.

5. The method according to any one of the preceding claims, wherein the model score is calculated using the following equation: in: Y is the biomarker score at the most recent time point, and is expressed as 0.4730. log[CA125] + 0.6531 log[CEA] + 0.2612 log[CYFRA21-1] + 0.9238 Calculate log[Pro-SFTPB]; µ is approximately 7.07; V is approximately 0.2672; B1 is approximately 0.767; B n Using equation (n) B1) / (n B1 + (1 – B1)) is calculated, where n is the total number of tests performed on the patient; and It is the average Y score calculated from biomarkers collected up to the most recent point in time.

6. The method according to any one of the preceding claims, wherein the predefined parameter is about 0.

63.

7. The method according to any one of claims 1-5, wherein the predefined parameter is about 0.

45.

8. The method according to any one of the preceding claims, wherein a model score greater than the predefined parameter is considered a positive test.

9. The method according to any one of claims 1-7, wherein a model score less than the predefined parameter is considered a negative test.

10. The method according to any one of the preceding claims, wherein scores of these biomarkers are collected at two to five time points.

11. The method of claim 10, wherein the scores of these biomarkers are collected at two time points.

12. The method of claim 10, wherein the scores of these biomarkers are collected at three time points.

13. The method of claim 10, wherein the scores of these biomarkers are collected at four time points.

14. The method of claim 10, wherein the scores of these biomarkers are collected at five time points.

15. The method according to any one of the preceding claims, wherein the levels of biomarkers CEA, CA125, CYFRA21-1 and Pro-SFTPB in a biological sample obtained from an individual are determined by immunoassay.

16. The method according to any one of the preceding claims, wherein each of the biomarkers CEA, CA125, CYFRA21-1 and Pro-SFTPB in the biological sample obtained from the individual generates a detectable signal.

17. The method of claim 16, wherein the detectable signal can be detected by spectroscopy.

18. The method of claim 17, wherein the spectroscopic method is selected from ultraviolet-visible spectroscopy, mass spectrometry, nuclear magnetic resonance (NMR) spectroscopy, proton NMR spectroscopy, nuclear magnetic resonance (NMR) spectroscopy, gas chromatography-mass spectrometry (GC-MS), liquid chromatography-mass spectrometry (LC-MS), correlation spectroscopy (COSY), nuclear Overhausen effect spectroscopy (NOESY), nuclear Overhausen effect spectroscopy in rotating coordinates (ROESY), time-of-flight LC-MS (LC-TOF-MS), liquid chromatography-tandem mass spectrometry (LC-MS / MS), and capillary electrophoresis-mass spectrometry.

19. The method of claim 18, wherein the spectroscopic method is mass spectrometry.

20. The method of claim 19, wherein the mass spectrometer is LC-TOF-MS.

21. The method according to any one of the preceding claims, wherein the lung cancer is early-stage (e.g., stage I or stage II).

22. The method according to any one of claims 1-20, wherein the lung cancer is advanced (e.g., stage III or IV).

23. The method according to any one of the preceding claims, wherein the individual has a smoking history of ≥10 pack-years.

24. The method according to any one of the preceding claims, wherein the individual is between 50 and 80 years old.

25. The method according to any one of the preceding claims, wherein the lead time prior to diagnosis of a lung cancer patient is greater than the lead time prior to diagnosis for different individual biomarkers, multiple biomarkers, groups, assays, algorithms, models, or any combination thereof.

26. The method of claim 25, wherein the pre-diagnosis period is greater than 1.03 years.

27. The method of claim 26, wherein the lead time prior to diagnosis is between 1.26 and 2.70 years.

28. The method according to any one of the preceding claims, wherein the patient is subsequently given further lung cancer screening or treatment.

29. The method of claim 28, wherein the screening is selected from endoscopic ultrasound, magnetic resonance imaging (MRI), and computed tomography (CT).

30. The method of claim 29, wherein the screening is performed annually.

31. The method of claim 29, wherein the screening is performed every six months.

32. The method of claim 28, wherein the treatment is selected from surgery, chemotherapy, immunotherapy, radiotherapy, targeted therapy, or a combination thereof.

Images