Prognosis evaluation method for biliary obstruction caused by pancreatic cancer

Abstract

The invention discloses a prognosis evaluation method for biliary obstruction caused by pancreatic cancer, which comprises the following steps: acquiring sample data, screening the sample data, adopting single-factor analysis and multi-factor analysis, selecting factors influencing lifetime from the sample data by using a COX proportional risk model, and evaluating the prognosis of the biliary obstruction caused by pancreatic cancer. According to sample data and factors influencing the lifetime, a model of a column diagram for lifetime prediction is established, a prediction model system is verified and compared by adopting a correction curve, a subject working curve and a decision curve, and the model established by the method has relatively good prediction accuracy; according to the method, the lifetime of the patient with the non-resectable pancreatic cancer with the biliary obstruction prepared to receive the biliary interventional operation can be evaluated and displayed in the form of a column graph, so that the method is intuitive, simple and clear, clinical doctors and the patient can operate the method conveniently, individualized prediction of the patient is performed, and an efficient method is provided for lifetime prediction of the patient with the pancreatic cancer with the biliary obstruction; and clinical interventional treatment and subsequent treatment decision are assisted.

Description

Technical Field

[0001] The present invention relates to the technical field of prognosis assessment of biliary obstruction, and in particular to a prognosis assessment method for biliary obstruction caused by pancreatic cancer. Background Art

[0002] Generally, the TNM staging guidelines can be used to assess whether pancreatic cancer patients who are diagnosed for the first time can be surgically resected, and the survival period can be roughly estimated for resectable pancreatic cancer ( Figure 6 ), it is worth noting that patients with biliary obstruction caused by pancreatic cancer or pancreatic cancer with biliary obstruction are very common in the clinical diagnosis and treatment process. Most patients with pancreatic cancer with biliary obstruction lose the opportunity for surgery, and biliary intervention is an effective means to relieve biliary obstruction, and can improve liver function for patients and strive for further treatment opportunities. Palliative treatment strategies are recommended for patients with an estimated survival period of less than 3 months, and appropriate anti-tumor treatment strategies can be considered for patients with an estimated survival period of more than 6 months.

[0003] At present, the pancreatic cancer survival prediction models that are discussed or used more frequently include CONUT score, PNI score and NPS score ( Figure 7 ) These scoring systems or models attempt to select prognostic risk factors from several clinical characteristics and use the risk factor scores to predict the prognosis of patients with surgically resected pancreatic cancer or advanced pancreatic cancer receiving chemotherapy. However, the characteristics selected by these models are not comprehensive enough and are biased towards nutritional and inflammatory characteristics, which means that their predictive efficacy needs to be further improved. In addition, there is currently no generally accepted prediction tool for the survival of patients with unresectable pancreatic cancer and biliary obstruction after interventional treatment (such as percutaneous transhepatic biliary drainage, percutaneous transhepatic biliary stent placement, etc.). Moreover, the above models have not been verified and compared for predicting the survival of patients with unresectable pancreatic cancer and biliary obstruction after interventional treatment.

[0004] Currently, interventional clinical practice faces the problem of a lack of specific and effective survival prediction tools for unresectable pancreatic cancer with biliary obstruction. Therefore, an improved technology is urgently needed to solve this problem in existing technologies. Summary of the invention

[0005] The purpose of the present invention is to provide a method with good predictive accuracy, which can evaluate the survival of patients with unresectable pancreatic cancer and biliary obstruction who are preparing to undergo biliary interventional surgery, and present it in the form of a nomogram, which is intuitive, concise and easy for clinicians and patients to operate, so as to perform individualized prediction of the prognosis evaluation method of biliary obstruction caused by pancreatic cancer for patients, so as to solve the problems raised in the above-mentioned background technology.

[0006] To achieve the above object, the present invention provides the following technical solution: a method for evaluating the prognosis of biliary obstruction caused by pancreatic cancer, wherein

[0007] Obtain sample data and screen the sample data;

[0008] Adopt univariate analysis and multivariate analysis, and use the COX proportional hazards model to select the factors affecting the survival period from the sample data;

[0009] Establish a nomogram model for survival prediction based on the sample data and the factors affecting the survival period;

[0010] Adopt calibration curves, receiver operating characteristic curves, and decision curves to verify and compare the prediction model system.

[0011] Preferably, the population from which the sample data is selected meets all of the following conditions:

[0012] a. At least 18 years old;

[0013] b. Clinically or histopathologically diagnosed with pancreatic cancer with biliary obstruction;

[0014] c. More than 1 month after surgery or a disease that cannot be resected due to extensive lesions, metastases, poor medical conditions, or refusal to undergo surgery;

[0015] d. Received percutaneous transhepatic biliary drainage catheter or biliary stent;

[0016] e. No history of any surgical operation within 1 month before the intervention;

[0017] f. No severe infection or organ failure before surgery;

[0018] g. The preoperative ECOG score is 0 - 3 points.

[0019] Preferably, the

[0020] The results of the univariate analysis show that the possible factors affecting the survival period include:

[0021] a. General parameters (gender);

[0022] b. Disease parameters (liver metastasis, M stage, preoperative infection, biliary obstruction location);

[0023] c. Immune parameters (percentage of blood lymphocytes);

[0024] d. Liver function parameters (serum albumin, serum total bilirubin, serum direct bilirubin);

[0025] e. Nutritional parameters (serum total cholesterol);

[0026] f. Renal function parameters (serum creatinine);

[0027] g. Electrolyte parameters (sodium in blood, calcium in blood);

[0028] h, tumor markers (CEA, CA199).

[0029] Preferably, the univariate analysis yields independent risk factors affecting survival. The factors affecting survival include liver metastasis, preoperative infection, biliary obstruction location, serum direct bilirubin, serum total cholesterol, and CA199, and these factors are independent risk factors affecting survival.

[0030] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0031] (1) Calibration curve fitting shows good consistency between the nomogram prediction of survival benefit and the actual survival situation;

[0032] (2) The nomogram model can divide the survival risks of patients with unresectable pancreatic cancer with biliary obstruction who undergo biliary intervention into a low-risk group and a high-risk group, thereby providing a decision-making basis for doctors to formulate treatment plans. The calibration curve, receiver operating characteristic curve, and decision curve are used to verify and compare the prediction model system;

[0033] (3) The model established by this method has good prediction accuracy, can evaluate the survival period of patients with unresectable pancreatic cancer with biliary obstruction who are prepared to undergo biliary intervention surgery, and is presented in the form of a nomogram, which is intuitive, concise, and convenient for clinical doctors and patients to operate for individual prediction of patients, providing an efficient method for predicting the survival period of patients with pancreatic cancer with biliary obstruction, and assisting in clinical intervention treatment and subsequent treatment decisions. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] Figure 1 It is a nomogram of the survival prediction model for patients with unresectable pancreatic cancer with biliary obstruction who undergo biliary intervention according to the present invention;

[0035] Figure 2 It is a schematic diagram of the calibration curve of the present invention for verifying the prediction accuracy of the scoring system of the nomogram;

[0036] Figure 3 It is a comparison data table of the area under the receiver operating characteristic curve data of the nomogram model of the present invention and other prediction model data;

[0037] Figure 4 It is a schematic diagram of the comparison of the decision curve of the nomogram model of the present invention with other prediction factors;

[0038] Figure 5 It is a schematic diagram of the Kaplan-Meier survival curve (A training group, B validation group) for dividing the survival risks of patients with unresectable pancreatic cancer with biliary obstruction who undergo biliary intervention into a low-risk group and a high-risk group according to the present invention;

[0039] Figure 6Schematic diagram of the TNM staging of pancreatic cancer and the survival period of resectable pancreatic cancer according to the present invention;

[0040] Figure 7 Schematic diagram of the prediction characteristics and risk stratification of the CONUT score, PNI score and NPS score according to the present invention. Detailed implementation manners

[0041] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0042] Please refer to Figures 1-7 , the present invention provides a technical solution: a prognostic evaluation method for biliary tract obstruction caused by pancreatic cancer,

[0043] First, sample data is obtained, and the sample data is screened. Univariate analysis and multivariate analysis are adopted, and the COX proportional hazards model is used to select the factors affecting the survival period from the sample data. According to the sample data and the factors affecting the survival period, a nomogram model for survival period prediction is established, and calibration curves, receiver operating characteristic curves, and decision curves are used to verify and compare the prediction model system.

[0044] Sample data is obtained, and the sample data is screened. The population of the sample data meets all the following conditions: at least 18 years old, clinically or histopathologically diagnosed with pancreatic cancer with biliary tract obstruction, more than 1 month after surgical operation or a disease that cannot be resected due to extensive lesions, metastasis, poor medical condition or refusal to accept surgery, receiving percutaneous transhepatic biliary drainage tube or biliary stent, no history of any surgical operation within 1 month before intervention, no severe infection or organ failure before surgery, and preoperative ECOG score of 0-3 points.

[0045] Univariate analysis and multivariate analysis are adopted. The results of univariate analysis show that the possible factors affecting the survival period include general parameters (gender), disease parameters (liver metastasis, M stage, preoperative infection, biliary tract obstruction location), immune parameters (percentage of blood lymphocytes), liver function parameters (serum albumin, total serum bilirubin, direct serum bilirubin), nutritional parameters (total serum cholesterol), renal function parameters (serum creatinine), electrolyte parameters (sodium in blood, calcium in blood), tumor markers (CEA, CA199), and then the COX proportional hazards model is used to select the factors affecting the survival period from the sample data.

[0046] Univariate analysis identified independent risk factors affecting survival. The factors affecting survival included liver metastasis, preoperative infection, biliary obstruction location, serum direct bilirubin, serum total cholesterol, and CA199. These factors were independent risk factors affecting survival.

[0047] Based on the sample data and the factors affecting survival, a nomogram model for predicting survival was established. The model established by this method can evaluate the survival of patients with unresectable pancreatic cancer with biliary obstruction who are prepared to undergo biliary intervention and is presented in the form of a nomogram, which is intuitive, concise, and convenient for clinicians and patients to operate for individualized prediction of patients.

[0048] The nomogram model had good predictive accuracy. The AUCs of the receiver operating characteristic curves at 1 month, 3 months, and 6 months in the training group were 0.933, 0.880, and 0.888 respectively, and the AUCs at 1 month, 3 months, and 6 months in the validation group were 0.801, 0.887, and 0.822 respectively. The calibration curve fitting showed good consistency between the predicted survival benefit of the nomogram and the actual survival situation.

[0049] The receiver operating characteristic curve of the nomogram model was superior to the following common prognostic prediction systems: NPS score, PNI score, CONUT score, and M stage. The decision curve of this model was also superior to the individual liver metastasis predictor, individual CA199 predictor, M stage combined with CA199 prediction, and liver metastasis combined with CA199 prediction.

[0050] The nomogram model can divide the survival risk of patients with unresectable pancreatic cancer with biliary obstruction undergoing biliary intervention into a low-risk group (≤115.2 points) and a high-risk group (>115.2 points), thus providing a decision-making basis for doctors to formulate treatment plans. The calibration curve, receiver operating characteristic curve, and decision curve were used to verify and compare the prediction model system.

[0051] A multi-dimensional clinical parameter prediction model for the survival of patients with unresectable pancreatic cancer with biliary obstruction was established. Through this model, the expected survival probability of patients with unresectable pancreatic cancer with biliary obstruction undergoing biliary intervention can be obtained.

[0052] According to the inclusion criteria for patients, all of the following conditions must be met: patients 18 years of age and older, clinically or histopathologically diagnosed with pancreatic cancer with biliary obstruction, more than 1 month after surgical operation or a disease that cannot be resected due to extensive lesions, metastasis, poor medical condition, or refusal to undergo surgery, percutaneous transhepatic biliary drainage tube or biliary stent placement. Patients meeting any of the following criteria need to be excluded: any surgical history within 1 month before intervention, severe infection or organ failure before surgery, preoperative ECOG score of 4, and sample data of such patients with incomplete clinical information or loss to follow-up.

[0053] The SPSS software (v26.0) was used to predict the survival probability by the COX proportional hazards regression model. Through univariate COX proportional hazards regression analysis, factors with a P-value < 0.1 were subjected to multivariate regression analysis. Using the forward LR method, factors with a P-value < 0.05 were retained in the model. The model was constructed based on the results of multivariate analysis in the training group and the COX proportional hazards regression model strategy.

[0054] The nomogram was drawn using R software (v3.4.4, www.r-project.org). The receiver operating characteristic curve and calibration curve were used to evaluate the predictive performance of the survival at 1, 3, and 6 months. When predicting whether the patients survived at 1, 3, and 6 months in the prediction samples, the areas under the receiver operating characteristic curve of the nomogram model were 0.933, 0.880, and 0.888 in the training group, and 0.801, 0.887, and 0.822 in the validation group, respectively. All of these values were greater than 0.75, showing good discrimination ability. The calibration curve indicated that the prediction results of the nomogram model were in good agreement with the actual survival results.

[0055] In addition to having a high specificity (0.727 - 0.958), the nomogram model also had good sensitivity (0.628 - 0.955). The net reclassification index (0.018–0.554) and integrated discrimination index (0.061–0.473) were both greater than 0, indicating that the nomogram model completely improved the predictive ability compared with other predictors. The decision curve analysis showed that the nomogram model had the greatest overall net benefit. Compared with M stage, liver metastasis, CA199, M stage combined with CA199, and liver metastasis combined with CA199, the nomogram model had better predictive performance in all aspects.

[0056] Decision curve analysis was used to evaluate the clinical application value of the nomogram by calculating the net benefit at different threshold probabilities. X-tile (v3.6.1) was used to determine the cut-off value to divide the patients into two groups with high or low survival periods. The Kaplan-Meier survival curve was used to evaluate the stratification ability of the high and low survival risk groups of patients, and the log-rank test was used to evaluate the statistical significance.

[0057] In the prediction of the 1-month mortality of 41 patients, the areas under the receiver operating characteristic curve of the NPS score, PNI score, CONUT score, M stage, liver metastasis, CA199, M stage combined with CA199, and liver metastasis combined with CA199 were 0.573, 0.523, 0.681, 0.687, 0.587, 0.573, 0.698, and 0.624, respectively. While the area under the receiver operating characteristic curve of the nomogram model in this invention was up to 0.829 at most.

[0058] Example 1

[0059] A total of 172 patients were included in this invention. There were 120 patients in the training group, including 78 males and 42 females, with an average age of 66.0 years ± 13.4 years. There were 52 patients in the validation group, including 31 males and 21 females, with an average age of 65.8 years ± 13.7 years. During the first biliary intervention, 57.6% (99 cases) of the patients received percutaneous transhepatic biliary stent placement, and 42.4% (73 cases) of the patients received percutaneous transhepatic biliary drainage.

[0060] In the training group, univariate analysis showed that factors related to survival included gender, liver metastasis, M stage, serum albumin, total serum bilirubin, direct serum bilirubin, total serum cholesterol, serum creatinine, serum sodium, serum calcium, lymphocyte ratio, CEA, CA199, preoperative infection, and biliary obstruction location (all P < 0.1). The optimal cut-off value of CA199 was determined to be 1000 U / mL, and the optimal cut-off value of CEA was 5 ng / mL. Among all samples, the median survival time of patients with CA199 ≥ 1000 U / mL was 69 days, and the median survival time of patients with CA199 < 1000 U / mL was 117 days (P = 0.003). The median survival time of patients with liver metastasis was 75 days, and the median survival time of patients without liver metastasis was 121 days (P < 0.001). The median survival time of patients with M stage metastasis was 80 days, and the median survival time of patients without M stage metastasis was 121 days (P = 0.003).

[0061] Preoperative infection was associated with poor prognosis. The median survival time of patients with preoperative infection was 41 days, and the median survival time of patients without preoperative infection was 90 days (P < 0.001). The survival time of patients with high-level obstruction was significantly lower than that of patients with low-level obstruction. The median survival time of patients with high-level obstruction was 60 days, and the median survival time of patients with low-level obstruction was 90 days (P = 0.004).

[0062] Multivariate COX proportional hazards regression analysis showed that the COX model was based on 6 clinical predictors, including direct serum bilirubin (HR (95% CI);

[0063] Multivariate COX proportional hazards regression analysis showed that the COX model was based on 6 clinical predictors, including

[0064] Direct serum bilirubin (HR (95% CI); 1.003 (1.001 - 1.005), P = 0.006),

[0065] Total serum cholesterol (HR (95% CI); 0.849 (0.764 - 0.943), P = 0.002),

[0066] Liver metastasis (HR (95% CI); 0.849 (0.764 - 0.943), P = 0.002),

[0067] CA199 (HR (95% CI); 2.211 (1.402 - 3.487), P = 0.001),

[0068] Preoperative infection (HR (95% CI); 3.474 (1.669 - 7.230), P = 0.001),

[0069] Obstruction location (HR (95% CI); 1.736 (1.098 - 2.746), P = 0.018),

[0070] A nomogram was established to visualize the model and facilitate its use.

[0071] According to the nomogram, the total score of 115.2 (risk probability of 0.72) was used as the cut-off value for risk stratification. Patients with unresectable pancreatic cancer with biliary obstruction were divided into a high-risk group (total score > 115.2 or risk probability > 0.72) and a low-risk group (total score ≤ 115.2 or risk probability ≤ 0.72).

[0072] Kaplan-Meier survival curves and log-rank tests showed a significant difference in overall survival between the high-risk group and the low-risk group (median survival in the training group: 37 days in the high-risk group and 137 days in the low-risk group, P < 0.001; median survival in the validation group: 41 days in the high-risk group and 120 days in the low-risk group, P < 0.001). Subgroup analysis showed that regardless of whether the patient had a history of surgical resection or received chemotherapy, the nomogram model had good performance in terms of the area under the receiver operating characteristic curve and risk stratification for predicting survival at 1, 3, and 6 months.

[0073] The model established by this method has good prediction accuracy, can evaluate the survival of patients with unresectable pancreatic cancer with biliary obstruction who are prepared to undergo biliary intervention surgery, and is presented in the form of a nomogram, which is intuitive, concise, and convenient for clinicians and patients to operate for individualized prediction of patients.

[0074] The present invention selects easily obtainable general information, clinical characteristics, and serological examination results from patients with unresectable pancreatic cancer with biliary obstruction who receive biliary intervention treatment, comprehensively integrates multi-dimensional clinical parameters such as the patient's basic situation, tumor characteristics, nutrition and inflammation indicators, liver and kidney function indicators, electrolyte internal environment indicators, and treatment methods, constructs a prediction model of a nomogram with strong intuitiveness, easy use, and high accuracy using a statistical model, verifies and compares the prediction efficacy of this model with TNM staging, CONUT score, PNI score, and NPS score, thereby providing an efficient method for predicting the survival of patients with pancreatic cancer with biliary obstruction to assist clinical interventional treatment and subsequent treatment decisions.

[0075] Although embodiments of the present invention have been shown and described, those of ordinary skill in the art will appreciate that various changes, modifications, substitutions, and variations can be made to these embodiments without departing from the principles and spirit of the present invention. The scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for evaluating the prognosis of biliary obstruction caused by pancreatic cancer, characterized in that: Said Obtain sample data and filter the sample data; Univariate analysis and multivariate analysis were performed, and the COX proportional hazard model was used to select factors affecting survival from the sample data; A nomogram model for predicting survival was established based on sample data and factors affecting survival; Calibration curves, receiver operating curves, and decision curves were used to verify and compare the prediction model system.

2. The method for evaluating the prognosis of biliary obstruction caused by pancreatic cancer according to claim 1, characterized in that: The sample data selection population meets all the following conditions: a. At least 18 years old; b. Clinical or histopathological diagnosis of pancreatic cancer with biliary obstruction; c. Disease that requires surgery for more than 1 month or cannot be removed due to extensive lesions, metastasis, poor medical condition or refusal to undergo surgery; d. Percutaneous transhepatic puncture for placement of biliary drainage tube or biliary stent; e. No history of any surgical operation within 1 month before intervention; f. No severe infection or organ failure before surgery; g. Preoperative ECOG score was 0-3 points.

3. The method for prognosis assessment of malignant biliary obstruction according to claim 1, characterized in that: The univariate analysis results showed that possible factors affecting survival include: a. General parameters (gender); b. Disease parameters (liver metastasis, M stage, preoperative infection, location of biliary obstruction); c. Immune parameters (blood lymphocyte percentage); d. Liver function parameters (serum albumin, serum total bilirubin, serum direct bilirubin); e. Nutritional parameters (serum total cholesterol); f. Renal function parameters (serum creatinine); g. Electrolyte parameters (serum sodium, serum calcium); h. Tumor markers (CEA, CA199).

4. The method for evaluating the prognosis of biliary obstruction caused by pancreatic cancer according to claim 1, characterized in that: The univariate analysis identified independent risk factors that affect survival, including liver metastasis, preoperative infection, location of biliary obstruction, serum direct bilirubin, serum total cholesterol, and CA199, which are independent risk factors that affect survival.

Images