Method for evaluating curative effect response of iodine therapy after total resection of papillary thyroid carcinoma by serum protein scoring system before radioiodine therapy
By using Olink proteomics detection and constructing a joint prediction model, serum protein biomarkers that may indicate poor response to radioactive iodine therapy in patients with papillary thyroid carcinoma were identified. This solved the problem of the inability to identify patients with poor response to treatment in the early stages of existing technologies, and achieved non-invasive, efficient assessment and accurate screening.
Patent Information
- Application Number
- CN202510901601.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-07-01
- Publication Date
- 2025-11-14
AI Technical Summary
Current technology cannot non-invasively assess the efficacy of iodine therapy after total resection of papillary thyroid carcinoma before radioactive iodine therapy, resulting in the inability to identify patients who may not respond well to treatment at an early stage.
Using Olink proteomics, serum samples from patients with papillary thyroid carcinoma before radioactive iodine therapy were measured after total resection. Serum protein levels were measured using 96 inflammation and 96 immunity panels. Differential proteins were screened as characteristic variables by combining LASSO regression analysis, independent samples t-test, and individual AUC ranking. A joint predictive model was constructed, and patients who may not respond well to radioactive iodine therapy were identified by dividing them into low-score and high-score groups.
This allows for the accurate screening of patients who may not respond well to radioactive iodine therapy before treatment, providing a window for early intervention, reducing the need for invasive testing, and improving the accuracy and convenience of assessment.
Smart Images
Figure CN120948801A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of medical technology, specifically to a method for evaluating the response to iodine therapy after total resection of papillary thyroid carcinoma using a serum protein scoring system prior to radioactive iodine therapy. Background Technology
[0002] According to the 2020 Global Cancer Incidence and Mortality Database compiled by the International Agency for Research on Cancer (IARC) of the World Health Organization, thyroid cancer is the most common malignant tumor of the endocrine system, ranking ninth in global cancer incidence. Papillary thyroid carcinoma (PTC) is the most important histopathological type, accounting for more than 85% of thyroid cancers. The most common first-line treatment option for PTC patients is radioactive iodine therapy (RAIT) after total thyroidectomy; however, the efficacy of RAIT varies from patient to patient. Different RAIT responses determine different subsequent clinical strategies and prognoses. The American Thyroid Association recommends long-term active follow-up for PTC patients, especially those at intermediate or high risk, after RAIT. The 2015 American Thyroid Association guidelines state that serological and imaging data 6–12 months after RAIT are used as indicators to assess the efficacy of treatment in PTC patients, dividing them into satisfactory response (ER) and non-satisfactory response (NER) groups.
[0003] However, almost all existing molecular classifications of PTC rely on tumor tissue removed by puncture or surgery. BRAF V600E Coexistence of TERT promoter mutations in PTC significantly increases tumor invasiveness, recurrence risk, and mortality. High Ki-67 expression is significantly associated with lymph node metastasis, TNM stage progression, and shortened disease-free survival in PTC. (Based on BRAF) V600E A comprehensive scoring system combining mutations, TERT promoter mutations, and the Ki-67 index can effectively differentiate between high-risk and low-risk PTC patients. However, these molecular classifications rely on immunohistochemistry or gene sequencing in clinical practice, and the tumor heterogeneity and limited biopsy depth in PTC also reduce the representativeness of biopsy samples to some extent. Furthermore, when PTC patients undergo surgery and are scheduled for RAIT, it is impossible to reassess the effectiveness of radioiodine therapy using biopsy samples.
[0004] Currently, there is no non-invasive method for assessing the efficacy of iodine therapy before RAIT. Therefore, it is crucial to have a reliable and convenient way to identify patients with poor RAIT response at an early stage. To this end, this invention proposes a method for assessing the response to iodine therapy after total resection of papillary thyroid carcinoma using a serum protein scoring system prior to radioactive iodine therapy. Summary of the Invention
[0005] In view of the problems existing in the prior art, the purpose of this invention is to provide a method for evaluating the response to iodine therapy after total resection of papillary thyroid carcinoma using a serum protein scoring system before radioactive iodine therapy. This method can accurately screen out patients who may not respond well to radioactive iodine therapy and proposes a patient stratification method and an early intervention time window.
[0006] To achieve the above objectives, the technical solution of this invention is as follows:
[0007] A method for evaluating the response to iodine therapy after total resection of papillary thyroid carcinoma using a serum protein scoring system prior to radioactive iodine therapy includes the following steps:
[0008] Step 1: Olink proteomics detection and analysis: Serum samples were collected from patients with papillary thyroid carcinoma before radioactive iodine therapy following total resection. 96 inflammation panels* and 96 Immunoassay Panel* measures serum protein levels;
[0009] Step 2: Based on the efficacy evaluation criteria, group the patients from Step 1 into those who are satisfied with the efficacy of RAIT and those who are not satisfied with the efficacy.
[0010] Step 3: Construction of the serum protein prediction scoring system before radioactive iodine therapy: Based on the protein data obtained in Step 1, differentially expressed proteins were selected as feature variables using three methods: LASSO regression analysis, independent samples t-test, and proteins with the highest single AUC values. These proteins were then used to construct a joint prediction model, namely the serum protein prediction scoring system before radioactive iodine therapy.
[0011] Step 4: Based on the optimal cutoff value in the joint prediction model established in Step 3, patients are divided into low-score and high-score groups.
[0012] Step 5: Analyze the percentage of patients with satisfactory and unsatisfactory treatment outcomes in the low-score and high-score groups obtained in Step 4, respectively.
[0013] Further, in step one, serum samples were extracted from patients with papillary thyroid carcinoma before radioactive iodine therapy following total resection. 96 inflammation panels* and The 96-immune panel* measures serum protein levels, and common proteins from the inflammation and immune panels are removed during statistical analysis.
[0014] Furthermore, the efficacy evaluation criteria in step two were based on the 2015 American Thyroid Association guidelines. Satisfactory efficacy was defined as negative imaging results 6–12 months after total thyroidectomy and radioactive iodine therapy, negative thyroglobulin antibodies, and suppressor thyroglobulin <0.2 ng / ml or stimulatory thyroglobulin <1 ng / ml. The remaining patients were defined as the poor efficacy group.
[0015] Furthermore, the Olink proteomics analysis in step one includes: using ggplot2 to perform GO and KEGG enrichment analysis on differentially expressed proteins.
[0016] Furthermore, the specific process of step three is as follows:
[0017] 3.1: The protein data obtained in step one were used to verify the intersection of three methods: LASSO regression analysis, independent samples t-test, and the protein with the highest single AUC value to determine the differentially expressed proteins as feature variables to be included in the joint diagnostic model.
[0018] 3.2: Then, plot the receiver operating characteristic curve for each differentially expressed protein in step 3.1, and evaluate the diagnostic efficacy of each individual protein based on its respective AUC value;
[0019] 3.3: By constructing a joint diagnostic model for all differentially expressed proteins in step 3.2 using binary logistic regression, the predicted probability value and the optimal cutoff value were calculated. It was found that the diagnostic efficacy of the joint prediction model was higher than that of the model constructed using a single protein. Therefore, the prediction model constructed using differentially expressed proteins can be used to establish a serum protein prediction scoring system.
[0020] Furthermore, in step four, based on the optimal cutoff value in the joint prediction model established in step three, the process of dividing patients into low-score and high-score groups is as follows:
[0021] If the predicted probability value obtained in step 3.3 is less than the optimal cutoff value, it is defined as the low predicted score group of the pre-treatment serum protein prediction scoring system; if the predicted probability value is greater than or equal to the optimal cutoff value, it is defined as the high predicted score group of the pre-treatment serum protein prediction scoring system.
[0022] Furthermore, the specific process of step five is as follows:
[0023] If more than 80% of patients in the low pre-treatment serum protein prediction score group had satisfactory treatment outcomes, and more than 90% of patients in the high pre-treatment serum protein prediction score group had poor treatment outcomes, it indicates that higher scores are associated with poorer treatment responses.
[0024] Compared with the prior art, the beneficial effects of the present invention are as follows:
[0025] 1) Compared with the prior art, the serum samples of the present invention can be easily obtained, causing little harm to patients. The equipment for detecting serum proteins is also widely used in various hospitals, and professional testing personnel are also widely employed.
[0026] 2) This invention is the first to discover biomarkers associated with poor response to iodine therapy after total resection of papillary thyroid carcinoma before radioactive iodine therapy. Specifically, by detecting the content of serum proteins, patients who may have poor response to radioactive iodine therapy can be screened relatively accurately, and a patient stratification method and a time window for early intervention are proposed. Attached Figure Description
[0027] Figure 1 (a) shows the LASSO coefficient pathway diagram for 141 proteins; (b) shows the cross-validation curves.
[0028] Figure 2 shows (a) the GO enrichment analysis diagram and (b) the KEGG enrichment analysis diagram.
[0029] Figure 3 (a) shows the ROC curve of the differential protein FASLG in predicting the response to radioactive iodine therapy;
[0030] (b) ROC curve of differential protein CXCL12 predicting response to radioiodine therapy; (c) ROC curve of differential protein HGF predicting response to radioiodine therapy.
[0031] Figure 4 ROC curves of the pre-RAIT serum protein combination in response to radioactive iodine therapy;
[0032] Figure 5 To calculate the ratio of pre-RAIT proteins using univariate logistic regression analysis;
[0033] Figure 6 The formula for calculating the pre-RAIT serum protein prediction score;
[0034] Figure 7 A comparison chart of treatment responses in patients with high and low PSP scores;
[0035] Figure 8 A diagram illustrating a pre-RAIT patient stratification management strategy based on pre-radioactive iodine therapy serum protein prediction scores. Figure 8 Created using the Figdraw online tool. Detailed Implementation
[0036] The present invention will be further described below with reference to the accompanying drawings and embodiments, but the scope of protection of the present invention is not limited to the scope described.
[0037] Example 1
[0038] Twenty-eight participants who visited the Department of Nuclear Medicine at the First Affiliated Hospital of Shanxi Medical University between January and August 2021 were included in this study. All participants underwent RAIT treatment prior to receiving the treatment. 99m TcO4-thyroid imaging and131 Whole-body scans revealed thyroid tissue contrast in some patients, with no significant difference in the amount of residual thyroid tissue among patients. This study was conducted in accordance with the principles of the Declaration of Helsinki and approved by the Ethics Committee of the First Affiliated Hospital of Shanxi Medical University (2021-K-K176).
[0039] The inclusion criteria were: (1) patients diagnosed with PTC; (2) patients who had undergone thyroidectomy; (3) patients who planned to receive RAIT treatment; and thyroid scans performed before RAIT. 131 I. Whole-body scintigraphy; (4) Before RAIT, levothyroxine was discontinued for 4 weeks and serum TSH increased to >30mIU / ml; (5) No glucocorticoid or immunosuppressive drug treatment was received within 4 months; (6) Levothyroxine was administered as replacement therapy after RAIT; (7) Peripheral blood samples were collected completely 1-2 days before RAIT; (8) The patient was willing to participate in this study and signed an informed consent form; (9) The patient promised to voluntarily accept and abide by this experimental protocol.
[0040] Exclusion criteria were: (1) patients with acute inflammatory diseases, autoimmune diseases, chronic inflammatory diseases, or other cancers and other diseases that may cause changes in immunity; and (2) patients who did not have a follow-up visit after RAIT.
[0041] Sixteen patients were in the satisfactory response (ER) group and 12 were in the non-satisfactory response (NER) group. The mean age of the participants was 41 years, and there were no statistically significant differences in age and BMI between the ER and NER groups (P>0.05). Women accounted for 71.43% of all cancer patients, with 87.50% in the ER group and 50.0% in the NER group.
[0042] A method for evaluating the response to iodine therapy after total resection of papillary thyroid carcinoma using a serum protein scoring system prior to radioactive iodine therapy includes the following steps:
[0043] Step 1: Serum samples were collected from the above 28 patients with papillary thyroid carcinoma before radioactive iodine therapy following total resection. Inflammation panel* and The immune panel* measured serum protein levels; proteins shared between the inflammation and immune panels were removed during statistical analysis, resulting in a total of 141 proteins included in the analysis. Specific information on these proteins is shown in the table below:
[0044]
[0045]
[0046]
[0047]
[0048]
[0049]
[0050]
[0051]
[0052]
[0053]
[0054] Olink proteomics assays utilize a dual-antibody system linked to DNA barcodes. The created double-stranded DNA "barcodes" are unique to the specific antigen and their number is proportional to the initial concentration of the target protein. Immediately after hybridization and extension, PCR amplification is performed, and the amplicon is finally quantified using microfluidic qPCR. Results are presented as log2-scaled normalized protein expression (NPX) values; higher NPX values indicate higher protein abundance.
[0055] As shown in Figure 2, GO and KEGG enrichment analyses of differentially expressed proteins were performed using ggplot2. GO enrichment analysis revealed that the differentially expressed proteins were significantly enriched in biological process pathways. No significant enrichment was observed in the KEGG pathway.
[0056] Step 2: Based on the efficacy evaluation criteria, group the patients from Step 1 into those who are satisfied with the efficacy of RAIT and those who are not satisfied with the efficacy.
[0057] In step two, the efficacy evaluation criteria were based on the 2015 American Thyroid Association guidelines. Satisfactory efficacy was defined as patients who had negative radiographic imaging, negative thyroglobulin antibody (TgAb), and suppressor thyroglobulin (Tg) <0.2 ng / ml or stimulatory thyroglobulin (Tg)(sTg) <1 ng / ml 6–12 months after total thyroidectomy and radioactive iodine therapy. The remaining patients were defined as the poor efficacy group.
[0058] Step 3: The protein data obtained in Step 1 is used to screen differentially expressed proteins as feature variables for constructing a joint prediction model by using three methods: LASSO regression analysis, independent samples t-test, and proteins with the highest single AUC values.
[0059] The specific process of step three is as follows:
[0060] 3.1: The 141 protein data obtained in Step 1 were analyzed using three methods: LASSO regression analysis (using five-fold cross-validation to determine the optimal regularization strength, selecting protein biomarkers with non-zero coefficients under this parameter), independent samples t-test (P<0.05), and validation of the intersection of individual AUC values to identify three differentially expressed proteins as feature variables for inclusion in the joint diagnostic model. Figure 1 It can be seen that the specific process is as follows:
[0061] 3.1.1: Using 141 proteins as independent variables, LASSO regression analysis was performed to generate a penalty function that reduces the number of variables in the regression model by the coefficients. When the standard error λ of the minimum distance is 0.247, three eigenvalues among the 141 independent variables are non-zero: FASLG, CXCL12, and HGF.
[0062] 3.1.2: Independent samples t-test was used to compare differentially expressed proteins between the ER and NER groups before radioactive iodine treatment. Differentially expressed proteins with P < 0.05 were selected as FASLG, CXCL12, and HGF.
[0063] 3.1.3: Calculate the AUC values of 141 proteins and select the top five proteins: FASLG, CXCL12, HGF, IL-10RB, and IL15;
[0064] 3.1.4: Jointly verify the differentially expressed proteins selected by the above three methods and take the intersection to finally determine FASLG, CXCL12 and HGF proteins as feature variables to be included in the joint diagnostic model;
[0065] 3.2: Then, plot the receiver operating characteristic curve for each differentially expressed protein in step 3.1, and evaluate the diagnostic efficacy of each individual protein based on its respective AUC value;
[0066] Depend on Figure 3 As shown, the optimal cutoff values for the three proteins were calculated using ROC curves. The optimal cutoff value for FASLG protein was defined as ≤7.41, AUC = 0.81 (95% CI: 0.648–0.977); the optimal cutoff value for CXCL12 protein was defined as ≤0.1, AUC = 0.73 (95% CI: 0.526–0.927); and the optimal cutoff value for HGF protein was defined as ≥9.22, AUC = 0.719 (95% CI: 0.5260–0.9115).
[0067] 3.3: A joint diagnostic model for all differentially expressed proteins from step 3.2 was constructed using binary logistic regression. Predicted probability values and optimal cutoff values were calculated. The results showed that the joint prediction model had higher diagnostic efficacy than models constructed using individual proteins. Figure 4As shown, the AUC of the combined diagnostic model was 0.885 (95% CI: 0.743–1.000), with an optimal cutoff value of 0.51. The combined diagnostic model based on multiple serum proteins has a higher predictive probability than the diagnostic model based on a single protein; therefore, the combined diagnostic model using multiple serum proteins is used to establish a serum protein prediction scoring system.
[0068] Step 4: The serum protein prediction scoring system established in Step 3 before radioactive iodine therapy will be used to re-divide the patients from Step 1 into two groups: those with satisfactory efficacy and those with poor efficacy.
[0069] The process of regrouping the patients from Step 1 in Step 4, based on the pre-radioactive iodine therapy serum protein prediction scoring system established in Step 3, is as follows:
[0070] Depend on Figure 5-8 It can be seen that if the predicted probability value obtained in step 3.3 is less than 0.51, it is defined as the low PSP predicted score group (PSP-L); if the predicted probability value is greater than or equal to 0.51, it is defined as the high PSP predicted score group (PSP-H).
[0071] In the PSP-L group, 93.8% of patients were in the ER group; in the PSP-H group, 83.3% were in the NER group. This indicates that higher scores are associated with poorer treatment response.
Claims
1. A method for evaluating the response to iodine therapy after total resection of papillary thyroid carcinoma using a serum protein scoring system prior to radioactive iodine therapy, characterized in that... Includes the following steps: Step 1: Olink proteomics detection and analysis: Serum samples were collected from patients with papillary thyroid carcinoma before radioactive iodine therapy following total resection. and Measure serum protein levels; Step 2: Based on the efficacy evaluation criteria, the patients from Step 1 are divided into a group satisfied with the efficacy of RAIT and a group with poor efficacy. Step 3: Construction of the serum protein prediction scoring system before radioactive iodine therapy: Based on the protein data obtained in Step 1, differentially expressed proteins were selected as feature variables using three methods: LASSO regression analysis, independent samples t-test, and proteins with the highest single AUC values. These proteins were then used to construct a joint prediction model, namely the serum protein prediction scoring system before radioactive iodine therapy. Step 4: Based on the optimal cutoff value in the joint prediction model established in Step 3, patients are divided into low-score and high-score groups. Step 5: Analyze the percentage of patients with satisfactory and unsatisfactory treatment outcomes in the low-score and high-score groups obtained in Step 4, respectively.
2. The method for evaluating the response to iodine therapy after total resection of papillary thyroid carcinoma using a serum protein scoring system prior to radioactive iodine therapy, as described in claim 1, is characterized in that... In step one, serum samples were collected from patients with papillary thyroid carcinoma before radioactive iodine therapy following total thyroidectomy. and Serum protein levels were measured, and common proteins in the inflammation and immunity panels were removed during statistical analysis.
3. The method for evaluating the response to iodine therapy after total resection of papillary thyroid carcinoma using a serum protein scoring system prior to radioactive iodine therapy, as described in claim 1, is characterized in that... In step two, the efficacy evaluation criteria were based on the 2015 American Thyroid Association guidelines. Satisfactory efficacy was defined as negative imaging results, negative thyroglobulin antibodies, and suppressor thyroglobulin < 0.2 ng / ml or stimulatory thyroglobulin < 1 ng / ml 6–12 months after total thyroidectomy and radioactive iodine therapy. The remaining patients were defined as the poor efficacy group.
4. The method for evaluating the response to iodine therapy after total resection of papillary thyroid carcinoma using a serum protein scoring system prior to radioactive iodine therapy, as described in claim 2, is characterized in that... Step one of the Olink proteomics analysis includes: using ggplot2 to perform GO and KEGG enrichment analysis on differentially expressed proteins.
5. The method for evaluating the response to iodine therapy after total resection of papillary thyroid carcinoma using a serum protein scoring system prior to radioactive iodine therapy, as described in claim 2, is characterized in that... The specific process of step three is as follows: 3.1: The protein data obtained in step one were used to verify the intersection of three methods: LASSO regression analysis, independent samples t-test, and the protein with the highest single AUC value to determine the differentially expressed proteins as feature variables to be included in the joint diagnostic model. 3.2: Then, plot the receiver operating characteristic curve for each differentially expressed protein in step 3.1, and evaluate the diagnostic efficacy of each individual protein based on its respective AUC value; 3.3: By constructing a joint diagnostic model for all differentially expressed proteins in step 3.2 using binary logistic regression, the predicted probability value and the optimal cutoff value were calculated. It was found that the diagnostic efficacy of the joint prediction model was higher than that of the model constructed using a single protein. Therefore, the prediction model constructed using differentially expressed proteins can be used to establish a serum protein prediction scoring system.
6. The method for evaluating the response to iodine therapy after total resection of papillary thyroid carcinoma using a serum protein scoring system prior to radioactive iodine therapy, as described in claim 5, is characterized in that... Step four involves classifying patients into low-score and high-score groups based on the optimal cutoff value established in step three: If the predicted probability value obtained in step 3.3 is less than the optimal cutoff value, it is defined as the low group of serum protein prediction score before radioactive iodine treatment; if the predicted probability value is greater than or equal to the optimal cutoff value, it is defined as the high group of serum protein prediction score before radioactive iodine treatment.
7. The method for evaluating the response to iodine therapy after total resection of papillary thyroid carcinoma using a serum protein scoring system prior to radioactive iodine therapy, as described in claim 1, is characterized in that... Step five involves the following steps: If more than 80% of patients in the low pre-treatment serum protein prediction score group had satisfactory treatment outcomes, and more than 90% of patients in the high pre-treatment serum protein prediction score group had poor treatment outcomes, it indicates that higher scores are associated with poorer treatment responses.