CN117741148A - Marker combination for predicting curative effect of immunotherapy and model construction method and application - Google Patents
Marker combination for predicting curative effect of immunotherapy and model construction method and application Download PDFInfo
- Publication number
- CN117741148A CN117741148A CN202410182047.8A CN202410182047A CN117741148A CN 117741148 A CN117741148 A CN 117741148A CN 202410182047 A CN202410182047 A CN 202410182047A CN 117741148 A CN117741148 A CN 117741148A
- Authority
- CN
- China
- Prior art keywords
- cells
- cell
- model
- positive rate
- panck
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000009169 immunotherapy Methods 0.000 title claims abstract description 40
- 239000003550 marker Substances 0.000 title claims abstract description 17
- 230000000694 effects Effects 0.000 title abstract description 46
- 238000010276 construction Methods 0.000 title abstract description 8
- 102000008096 B7-H1 Antigen Human genes 0.000 claims abstract description 73
- 108010074708 B7-H1 Antigen Proteins 0.000 claims abstract description 73
- 206010028980 Neoplasm Diseases 0.000 claims abstract description 57
- 102000001398 Granzyme Human genes 0.000 claims abstract description 22
- 108060005986 Granzyme Proteins 0.000 claims abstract description 22
- 101000934372 Homo sapiens Macrosialin Proteins 0.000 claims abstract description 18
- 102100025136 Macrosialin Human genes 0.000 claims abstract description 18
- 230000001225 therapeutic effect Effects 0.000 claims abstract description 9
- 101000946843 Homo sapiens T-cell surface glycoprotein CD8 alpha chain Proteins 0.000 claims abstract description 5
- 102100034922 T-cell surface glycoprotein CD8 alpha chain Human genes 0.000 claims abstract description 5
- 210000004027 cell Anatomy 0.000 claims description 140
- 238000000034 method Methods 0.000 claims description 34
- 210000001519 tissue Anatomy 0.000 claims description 31
- 208000002154 non-small cell lung carcinoma Diseases 0.000 claims description 26
- 208000029729 tumor suppressor gene on chromosome 11 Diseases 0.000 claims description 26
- 238000004458 analytical method Methods 0.000 claims description 21
- 230000014509 gene expression Effects 0.000 claims description 17
- 238000010186 staining Methods 0.000 claims description 17
- 238000004364 calculation method Methods 0.000 claims description 16
- 238000000116 DAPI staining Methods 0.000 claims description 12
- FWBHETKCLVMNFS-UHFFFAOYSA-N 4',6-Diamino-2-phenylindol Chemical compound C1=CC(C(=N)N)=CC=C1C1=CC2=CC=C(C(N)=N)C=C2N1 FWBHETKCLVMNFS-UHFFFAOYSA-N 0.000 claims description 11
- 230000000007 visual effect Effects 0.000 claims description 8
- 101000738771 Homo sapiens Receptor-type tyrosine-protein phosphatase C Proteins 0.000 claims description 7
- 102100037422 Receptor-type tyrosine-protein phosphatase C Human genes 0.000 claims description 7
- 239000000427 antigen Substances 0.000 claims description 7
- 102000036639 antigens Human genes 0.000 claims description 7
- 108091007433 antigens Proteins 0.000 claims description 7
- 238000007789 sealing Methods 0.000 claims description 5
- 229940125644 antibody drug Drugs 0.000 claims description 4
- 210000003855 cell nucleus Anatomy 0.000 claims description 4
- 238000004043 dyeing Methods 0.000 claims description 4
- 238000011534 incubation Methods 0.000 claims description 4
- 210000004369 blood Anatomy 0.000 claims description 3
- 239000008280 blood Substances 0.000 claims description 3
- 239000000470 constituent Substances 0.000 claims description 3
- 210000000805 cytoplasm Anatomy 0.000 claims description 3
- 210000000981 epithelium Anatomy 0.000 claims description 3
- 239000007850 fluorescent dye Substances 0.000 claims description 3
- 230000001024 immunotherapeutic effect Effects 0.000 claims description 3
- 210000003205 muscle Anatomy 0.000 claims description 3
- 238000012758 nuclear staining Methods 0.000 claims description 3
- 230000000887 hydrating effect Effects 0.000 claims description 2
- 230000008439 repair process Effects 0.000 claims description 2
- 238000002651 drug therapy Methods 0.000 claims 1
- 238000001514 detection method Methods 0.000 abstract description 23
- 238000003364 immunohistochemistry Methods 0.000 abstract description 20
- 230000008901 benefit Effects 0.000 abstract description 16
- 238000005516 engineering process Methods 0.000 abstract description 8
- 238000010166 immunofluorescence Methods 0.000 abstract description 4
- 230000001744 histochemical effect Effects 0.000 abstract 1
- 230000004083 survival effect Effects 0.000 description 28
- 101710089372 Programmed cell death protein 1 Proteins 0.000 description 16
- 239000000090 biomarker Substances 0.000 description 16
- 238000011282 treatment Methods 0.000 description 15
- 210000002865 immune cell Anatomy 0.000 description 13
- 230000004069 differentiation Effects 0.000 description 9
- 238000012216 screening Methods 0.000 description 9
- 238000012549 training Methods 0.000 description 9
- 230000000052 comparative effect Effects 0.000 description 8
- 210000003171 tumor-infiltrating lymphocyte Anatomy 0.000 description 8
- 230000000875 corresponding effect Effects 0.000 description 7
- 108090000623 proteins and genes Proteins 0.000 description 7
- 102100022005 B-lymphocyte antigen CD20 Human genes 0.000 description 6
- 101000897405 Homo sapiens B-lymphocyte antigen CD20 Proteins 0.000 description 6
- KHGNFPUMBJSZSM-UHFFFAOYSA-N Perforine Natural products COC1=C2CCC(O)C(CCC(C)(C)O)(OC)C2=NC2=C1C=CO2 KHGNFPUMBJSZSM-UHFFFAOYSA-N 0.000 description 6
- 201000011510 cancer Diseases 0.000 description 6
- 230000008045 co-localization Effects 0.000 description 6
- 238000011065 in-situ storage Methods 0.000 description 6
- 229930192851 perforin Natural products 0.000 description 6
- 102000004169 proteins and genes Human genes 0.000 description 6
- 102100034458 Hepatitis A virus cellular receptor 2 Human genes 0.000 description 5
- 101710083479 Hepatitis A virus cellular receptor 2 homolog Proteins 0.000 description 5
- 101000831007 Homo sapiens T-cell immunoreceptor with Ig and ITIM domains Proteins 0.000 description 5
- 102000017578 LAG3 Human genes 0.000 description 5
- 101150030213 Lag3 gene Proteins 0.000 description 5
- 229940126547 T-cell immunoglobulin mucin-3 Drugs 0.000 description 5
- 102100024834 T-cell immunoreceptor with Ig and ITIM domains Human genes 0.000 description 5
- 230000008569 process Effects 0.000 description 5
- 238000010200 validation analysis Methods 0.000 description 5
- 102100023990 60S ribosomal protein L17 Human genes 0.000 description 4
- 101000874179 Homo sapiens Syndecan-1 Proteins 0.000 description 4
- 102100035721 Syndecan-1 Human genes 0.000 description 4
- 238000003556 assay Methods 0.000 description 4
- 238000013461 design Methods 0.000 description 4
- 238000003745 diagnosis Methods 0.000 description 4
- 239000003814 drug Substances 0.000 description 4
- 210000002540 macrophage Anatomy 0.000 description 4
- 230000001537 neural effect Effects 0.000 description 4
- 238000003908 quality control method Methods 0.000 description 4
- 108091032973 (ribonucleotides)n+m Proteins 0.000 description 3
- 102000037982 Immune checkpoint proteins Human genes 0.000 description 3
- 108091008036 Immune checkpoint proteins Proteins 0.000 description 3
- 206010058467 Lung neoplasm malignant Diseases 0.000 description 3
- 238000003559 RNA-seq method Methods 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 229940079593 drug Drugs 0.000 description 3
- 238000011156 evaluation Methods 0.000 description 3
- 238000002474 experimental method Methods 0.000 description 3
- 239000011521 glass Substances 0.000 description 3
- 230000028993 immune response Effects 0.000 description 3
- 201000005202 lung cancer Diseases 0.000 description 3
- 208000020816 lung neoplasm Diseases 0.000 description 3
- 238000010801 machine learning Methods 0.000 description 3
- 210000003071 memory t lymphocyte Anatomy 0.000 description 3
- 238000012545 processing Methods 0.000 description 3
- 238000004393 prognosis Methods 0.000 description 3
- 238000002331 protein detection Methods 0.000 description 3
- 230000004044 response Effects 0.000 description 3
- 238000002560 therapeutic procedure Methods 0.000 description 3
- 210000004881 tumor cell Anatomy 0.000 description 3
- 238000012795 verification Methods 0.000 description 3
- LFQSCWFLJHTTHZ-UHFFFAOYSA-N Ethanol Chemical compound CCO LFQSCWFLJHTTHZ-UHFFFAOYSA-N 0.000 description 2
- WSFSSNUMVMOOMR-UHFFFAOYSA-N Formaldehyde Chemical compound O=C WSFSSNUMVMOOMR-UHFFFAOYSA-N 0.000 description 2
- 229940076838 Immune checkpoint inhibitor Drugs 0.000 description 2
- 102000037984 Inhibitory immune checkpoint proteins Human genes 0.000 description 2
- 108091008026 Inhibitory immune checkpoint proteins Proteins 0.000 description 2
- 210000001744 T-lymphocyte Anatomy 0.000 description 2
- 230000002159 abnormal effect Effects 0.000 description 2
- 230000005856 abnormality Effects 0.000 description 2
- 238000009175 antibody therapy Methods 0.000 description 2
- 210000003719 b-lymphocyte Anatomy 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 2
- 238000006243 chemical reaction Methods 0.000 description 2
- 239000003153 chemical reaction reagent Substances 0.000 description 2
- 238000002512 chemotherapy Methods 0.000 description 2
- 210000001151 cytotoxic T lymphocyte Anatomy 0.000 description 2
- 201000010099 disease Diseases 0.000 description 2
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 2
- 239000012274 immune-checkpoint protein inhibitor Substances 0.000 description 2
- 239000007788 liquid Substances 0.000 description 2
- 201000001441 melanoma Diseases 0.000 description 2
- 238000010197 meta-analysis Methods 0.000 description 2
- 210000004940 nucleus Anatomy 0.000 description 2
- 238000007781 pre-processing Methods 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 238000012163 sequencing technique Methods 0.000 description 2
- 239000000243 solution Substances 0.000 description 2
- 238000013517 stratification Methods 0.000 description 2
- 239000012224 working solution Substances 0.000 description 2
- 206010005003 Bladder cancer Diseases 0.000 description 1
- 206010064571 Gene mutation Diseases 0.000 description 1
- 108010001336 Horseradish Peroxidase Proteins 0.000 description 1
- 102000011782 Keratins Human genes 0.000 description 1
- 108010076876 Keratins Proteins 0.000 description 1
- 102000018697 Membrane Proteins Human genes 0.000 description 1
- 108010052285 Membrane Proteins Proteins 0.000 description 1
- 206010061309 Neoplasm progression Diseases 0.000 description 1
- CTQNGGLPUBDAKN-UHFFFAOYSA-N O-Xylene Chemical compound CC1=CC=CC=C1C CTQNGGLPUBDAKN-UHFFFAOYSA-N 0.000 description 1
- 102000004503 Perforin Human genes 0.000 description 1
- 108010056995 Perforin Proteins 0.000 description 1
- 208000005718 Stomach Neoplasms Diseases 0.000 description 1
- 101710194092 Thiamine-phosphate synthase 1 Proteins 0.000 description 1
- 208000003721 Triple Negative Breast Neoplasms Diseases 0.000 description 1
- 208000007097 Urinary Bladder Neoplasms Diseases 0.000 description 1
- 238000009098 adjuvant therapy Methods 0.000 description 1
- 230000003321 amplification Effects 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000000903 blocking effect Effects 0.000 description 1
- 238000009835 boiling Methods 0.000 description 1
- 238000009096 combination chemotherapy Methods 0.000 description 1
- 238000012790 confirmation Methods 0.000 description 1
- 238000001816 cooling Methods 0.000 description 1
- 230000002596 correlated effect Effects 0.000 description 1
- 231100000433 cytotoxic Toxicity 0.000 description 1
- 230000001472 cytotoxic effect Effects 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 238000007418 data mining Methods 0.000 description 1
- 230000018044 dehydration Effects 0.000 description 1
- 238000006297 dehydration reaction Methods 0.000 description 1
- 238000007598 dipping method Methods 0.000 description 1
- 239000006185 dispersion Substances 0.000 description 1
- 239000000975 dye Substances 0.000 description 1
- 210000003162 effector t lymphocyte Anatomy 0.000 description 1
- 238000011261 first line immunotherapy Methods 0.000 description 1
- 238000009472 formulation Methods 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 206010017758 gastric cancer Diseases 0.000 description 1
- 210000002443 helper t lymphocyte Anatomy 0.000 description 1
- 230000036571 hydration Effects 0.000 description 1
- 238000006703 hydration reaction Methods 0.000 description 1
- 238000010191 image analysis Methods 0.000 description 1
- 238000003709 image segmentation Methods 0.000 description 1
- 238000003384 imaging method Methods 0.000 description 1
- 230000005746 immune checkpoint blockade Effects 0.000 description 1
- 230000036039 immunity Effects 0.000 description 1
- 238000013115 immunohistochemical detection Methods 0.000 description 1
- 230000008595 infiltration Effects 0.000 description 1
- 238000001764 infiltration Methods 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 238000002372 labelling Methods 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 108020004999 messenger RNA Proteins 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
- 230000035772 mutation Effects 0.000 description 1
- 238000009099 neoadjuvant therapy Methods 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 238000003199 nucleic acid amplification method Methods 0.000 description 1
- 239000012188 paraffin wax Substances 0.000 description 1
- 230000001575 pathological effect Effects 0.000 description 1
- 230000037361 pathway Effects 0.000 description 1
- 210000004180 plasmocyte Anatomy 0.000 description 1
- 238000002203 pretreatment Methods 0.000 description 1
- 238000010791 quenching Methods 0.000 description 1
- 210000003289 regulatory T cell Anatomy 0.000 description 1
- 238000012552 review Methods 0.000 description 1
- 230000035945 sensitivity Effects 0.000 description 1
- 238000009097 single-agent therapy Methods 0.000 description 1
- 238000000638 solvent extraction Methods 0.000 description 1
- 241000894007 species Species 0.000 description 1
- 238000011272 standard treatment Methods 0.000 description 1
- 201000011549 stomach cancer Diseases 0.000 description 1
- 230000009897 systematic effect Effects 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 230000004797 therapeutic response Effects 0.000 description 1
- 238000013519 translation Methods 0.000 description 1
- 208000022679 triple-negative breast carcinoma Diseases 0.000 description 1
- 230000005748 tumor development Effects 0.000 description 1
- 230000005751 tumor progression Effects 0.000 description 1
- 201000005112 urinary bladder cancer Diseases 0.000 description 1
- 238000005406 washing Methods 0.000 description 1
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 1
- 239000001993 wax Substances 0.000 description 1
- 230000037303 wrinkles Effects 0.000 description 1
- 239000008096 xylene Substances 0.000 description 1
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A90/00—Technologies having an indirect contribution to adaptation to climate change
- Y02A90/10—Information and communication technologies [ICT] supporting adaptation to climate change, e.g. for weather forecasting or climate simulation
Landscapes
- Investigating Or Analysing Biological Materials (AREA)
Abstract
The invention provides a marker combination for predicting an immune therapeutic effect, a model construction method and application, and belongs to the field of immune histochemical detection. The marker combination of the invention consists of the following markers: granzyme B, CD68, CD8, PD-L1, PANCK, foxp3 and CD45RO. The invention utilizes the advantages of multicolor immunohistochemistry/immunofluorescence compared with the traditional PD-L1 IHC technology to comprehensively analyze the tumor immune microenvironment of a patient, and obtains a more accurate curative effect prediction model aiming at immunotherapy.
Description
Technical Field
The invention belongs to the field of immunohistochemical detection, and particularly relates to a marker combination for predicting an curative effect of immunotherapy, a model construction method and application.
Background
Immune Checkpoint Inhibitors (ICIs) have become the basis for a variety of tumor treatments as a major component of immunotherapy. anti-PD-1 mab reactivates immune responses to tumor cells by blocking the interaction of PD-L1 and PD-1, and several studies have demonstrated that it can bring higher clinical benefit to patients than traditional chemotherapy. Nonetheless, the Objective Response Rate (ORR) of anti-PD-1 mab was only about 24% (95% CI,21% -28%). Thus, there is a need to develop methods and products with robust predictive capabilities for anti-PD-1 mab, thereby better guiding patient stratification and clinical decisions.
Methods for predicting therapeutic efficacy against PD-1 mab therapy mainly include assessing PD-L1 protein expression, tumor Mutation Burden (TMB), and Gene Expression Profile (GEP):
the detection of PD-L1 protein expression by PD-L1 Immunohistochemistry (IHC) is the most widely applied and most evidence immunotherapy curative effect prediction method at present, and can be used as the accompanying or supplementary diagnosis of patients with non-small cell lung cancer (NSCLC), melanoma, bladder cancer and gastric cancer. Although PD-L1 expression has been shown to be associated with therapeutic response to certain tumor types, there are a number of limitations to PD-L1 IHC as a diagnostic assay: (1) Each slice can only detect a single target and can provide very limited information. (2) The heterogeneity of PD-L1 expression within tumors makes local visual field analysis limiting. (3) Different PD-L1 detection products have the differences of antibodies, detection environments, interpretation methods and PD-L1 positive thresholds, and the problem of standardized detection and interpretation is faced at the present stage. (4) Reproducibility of the test results between different pathologists is problematic. (5) more importantly: PD-L1 expression is only one of the Tumor Immune Microenvironment (TIME) factors affecting tumor progression, and the detection of PD-L1 alone lacks an overall view of the tumor immune microenvironment. Furthermore, traditional IHC detection does not allow analysis of the spatial distribution of different cells, whereas the spatial distribution of cells in TIME is considered one of the main causes of tumor spatial heterogeneity and is related to prognosis of tumors;
in clinical practice, not all PD-L1 highly expressed patients benefit from immunotherapy on the one hand, and some PD-L1 lowexpressed patients still benefit from immunotherapy on the other hand, PD-L1 IHC detection has limited predictions of the efficacy of immunotherapy. Thus, in addition to PD-L1 IHC detection, other markers predictive of efficacy of immune checkpoint therapeutics are also constantly emerging. TMB is one of them and has moved from clinical studies to clinical applications. However, TMB also has many disadvantages as a predictive marker: (1) The amount of TMB does not accurately predict the immune response because it is not in strict proportion to the amount of gene mutations that can elicit an immune response; (2) the TMB level of different types of tumors varies greatly; (3) Age has a large influence on TMB expression levels, and the relationship between the two varies depending on the type of disease; (4) The detection of TMB is far more costly than PD-L1 detection by immunohistochemistry due to the reliance on gene sequencing; (5) the detection standards of different platforms are not uniform;
in addition, several studies suggest that GEP also has a certain predictive value of efficacy for immunotherapy. However, GEPs based on general transcriptome sequencing (Bulk RNA-seq) also suffer from disadvantages: (1) The expression of RNA in tumor has heterogeneity, and Bulk RNA-Seq cannot reflect the heterogeneity due to the lack of tissue in-situ information; (2) An abnormality in the RNA level does not necessarily lead to an abnormality in the function of the protein, and thus detection in the protein level may still be required; (3) RNA sequencing analysis is not a widely accepted reference standard in clinical practice.
In addition to the above biomarkers, more and more data suggest that immune cells and other biomarkers on the surface in TIME are also of great importance in guiding drug selection and prognosis evaluation in tumor patients. For example, researchers such as PD-1 discoverers show professor (d) that divide TIME into four different types based on the expression of PD-L1 in TIME and the infiltration of Tumor Infiltrating Lymphocytes (TILs): type I: PD-L1 - /TILs - The method comprises the steps of carrying out a first treatment on the surface of the Type II: PD-L1 + /TILs + The method comprises the steps of carrying out a first treatment on the surface of the Type III: PD-L1 - /TILs + The method comprises the steps of carrying out a first treatment on the surface of the Type IV: PD-L1 + /TILs - . These four different TIME types are closely related to the therapeutic efficacy of immunotherapy. Also, studies have shown that co-localization of macrophage surface protein CD68 with PD-L1 is not only closely related to the efficacy of NSCLC immunotherapy, but also to the prognosis of triple negative breast cancer. In addition, several immunotherapy cohort analyses showed that B cell-related biomarkers correlated with melanoma immunotherapy efficacy and NSCLC immunotherapy efficacy.
The above-mentioned research reports demonstrate that different immune cell subsets in TIME are closely related to the effects of immunotherapy, and there is a need for a method that can detect and evaluate complex immune cell subsets and their correlations in TIME, thereby more accurately guiding patient stratification of immunotherapy.
Multiple immunohistochemistry/immunofluorescence (mhic/IF) is one such new technique. Several studies have shown that mIHC/IF is superior to other methods in judging the efficacy of anti-PD-1/PD-L1 immunotherapy. There are 56 prior art systems that analyze the correlation of different detection methods (PD-L1 IHC, TMB, GEP, mIHC/IF) with anti-PD-1/PD-L1 immunotherapeutic responses. The study shows that the mIHC/IF method has higher efficacy prediction value for tumor immunotherapy compared with PD-L1 IHC, TMB and GEP evaluation methods. The predicted evaluation index AUC values of the PD-L1 IHC, TMB, GEP on the curative effect of the tumor immunotherapy are respectively 0.650, 0.688 and 0.650, and the AUC value of mIHC/IF obtained by meta-analysis is as high as 0.790, which is obviously higher than that of the other three methods. In addition, the efficacy prediction assessment index AUC of the combination of the three methods of PD-L1 IHC, TMB, GEP was 0.736, which was also lower than the single index result of mIHC/IF (Lu, S., et al Comparison of Biomarker Modalities for Predicting Response to PD-1/PD-L1 Checkpoint Blockade: A Systematic Review and Meta-analysis, JAMA Oncol, 2019.5 (8): p.1195-1204).
Monotherapy and combination chemotherapy with anti-PD-1/PD-L1 immunotherapy as the core have become standard treatment regimens for NSCLC and gradually entered into perioperative neoadjuvant and adjuvant treatment regimens for early stage lung cancer. Numerous mhic/IF-based studies have shown that multiple immune cell subsets in TIME have significant predictive value of efficacy in NSCLC immunotherapy: CD8 + T Cell and spatial Immune heterogeneity (Lopez de Rodas, M., et al, role of Tumor infiltrating lymphocytes and spatial Immune heterogeneity in sensitivity to PD-1 axis blockers in non-small Cell lung Cancer J Immunother Cancer, 2022.10 (6)), co-localization of PD-L1 and CD68 (Liu, Y., et al, immune Cell PD-L1 Colocalizes with Macrophages and Is Associated with Outcome in PD-1 Pathway Blockade Therapy. Clin Cancer Res, 2020.26 (4): p.970-977), co-localization of PD-1 and CD8 (Mazzarski, G., et al, low PD-1 Expression in Cytotoxic CD8 (+) Tumor-Infiltrating Lymphocyte)s Confers an Immune-Privileged Tissue Microenvironment in NSCLC with a Prognostic and Predictive value, clin Cancer Res, 2018.24 (2): p.407-419) and the like are biomarkers for predicting the efficacy of immunotherapy of lung Cancer. Not only lung cancer, but also other cancer species have been shown to have various immune cell subsets associated with therapeutic efficacy of immunotherapy. Thus, accurate guidance of anti-PD-1 therapy using mhic/IF has urgent clinical need for tumor TIME analysis.
Disclosure of Invention
In order to solve the problems, the invention provides a model and a method for predicting the curative effect of anti-PD 1 antibody medicines on non-small cell lung cancer. The mIHC/IF technology related by the invention utilizes horseradish peroxidase to carry out high-density in-situ labeling on target protein and carries out Tyramide Signal Amplification (TSA), thereby combining immunofluorescence and the traditional IHC technology, and enabling quantitative detection on complex immune cell subsets in TIME to be possible. Compared with the traditional IHC technology which can only detect one target molecule, the mIHC/IF technology can determine the expression, abundance and spatial position of a plurality of target proteins on the same tissue sample, provides a more comprehensive and detailed immunohistological image, and helps to comprehensively understand the effect of TIME in tumor development and treatment reaction.
The invention firstly provides a marker combination screen for predicting the curative effect of immunotherapy, which is selected from the following markers:
an immune cell subtype marker comprising: CD3, CD4, CD8, foxp3, CD68, CD103, CD45RO, CD20, CD138, granzyme B (Granzyme B)/Perforin (Perforin);
immune checkpoint/therapeutic targets, including PD-L1, PD-1, LAG-3, TIM-3, TIGIT.
Preferably, the marker combination is predicted from the following 5 panels:
panel 1: CD8, CD103, granzyme B, perforin, PANCK (cytokeratin);
panel 2: CD8, LAG-3, TIM-3, TIGIT, PANCK;
panel 3: CD8, PD-L1, PD-1, CD68, PANCK;
panel 4: CD4, CD8, CD45RO, foxp3, PANCK;
panel 5: CD3, CD8, CD20, CD138, PANCK.
The final screened marker combination of the invention consists of the following markers: granzyme B, CD68, CD8, PD-L1, PANCK, foxp3 and CD45RO.
The invention also provides application of the marker combination in constructing an immune treatment effect prediction model.
Based on the above, the invention also provides a therapeutic effect prediction model for immunotherapy constructed by the marker combination.
Specifically, monolithic PD-L1 is employed + PANCK + Cell positive rate, overall CD45RO + Cell positive rate, interstitial region CD8 + Cell positive rate, interstitial region Granzyme B + Cell positive rate, tumor zone CD68 + PD-L1 + Cell positive rate and Foxp3 within 100 microns + Cell to PANCK + The shortest distance of cells was constructed.
The model is used for calculating the immune score, and the immune score calculation mode is as follows:
(0.20-0.45) × (interstitial region Granzyme B + Positive rate of cells) + (0.40-0.80) × (tumor area CD 68) + PD-L1 + Positive rate of cells) + (0.10-0.20) × (interstitial region CD 8) + Cell positive) + (0.05-0.15) × (Whole CD45 RO) + Positive rate of cells) + (0.05-0.15) × (Whole area PANCK + PD-L1 + Cell positive) + (0.70-0.90) × (Foxp 3 within 100 microns) + Cell to PANCK + Shortest distance of cells).
The predictive cut-off for the immune score was 0.5.
Preferably, the immune score calculation method is as follows:
0.42× (interstitial region Granzyme B + Positive rate of cells) +0.63× (tumor zone CD 68) + PD-L1 + Positive rate of cells) +0.20× (interstitial region CD8 + Cell positive) +0.10× (global CD45 RO) + Positive rate of cells) +0.10× (Whole area PANCK + PD-L1 + Cell positive) +0.81× (Foxp 3 within 100 microns) + Cell to PANCK + Shortest distance of cells).
The invention also provides a method for constructing an immune treatment effect prediction model by the marker combination, which comprises the following steps:
step (1): FFPE specimen mIHC is dyed;
step (2): scanning and film reading;
step (3): analyzing data;
step (4): and (5) constructing an immune treatment efficacy prediction model.
Specifically, the step (1) includes: baking slices; dewaxing; hydrating; antigen retrieval; closing; incubation with primary antibody; incubating the secondary antibody; amplifying the signal by fluorescent staining; repeating antigen repair after each round of antibody staining, and carrying out the next round of antibody staining; sealing; and reading the film.
Specifically, the step (2) includes: after scanning, clearly and completely dividing analysis areas of the images, wherein the dividing rules are as follows:
1) Areas with no/few DAPI nuclear staining: no DAPI staining is carried out, no matter whether other channels are stained or not, all the channels are removed; small/sporadic DAPI staining, if blood and muscle tissue, total rejection;
2) Non-cellular constituent rejection with DAPI staining;
3) Folding area: removing a folding area caused by tissue flaking partially in the dyeing scanning image;
4) Removing scattered areas which are free from the whole sample;
5) Normal epithelial tissue rejection.
Specifically, the step (3) includes:
A. tissue splitting: tumor tissue and tumor stroma resolution;
B. cell resolution: cell nucleus and cytoplasm resolution;
C. calculating the characteristics of the cell positive rate;
D. positive cell positioning, calculating the distance between positive cells, and analyzing the spatial tissue distribution characteristics of cells.
Further specifically, in the A, resolution is performed by DAPI staining and PANCK staining;
further specifically, in the step B, resolution is performed through the expression intensity of DAPI;
further specifically, in the C, positive cells of different subtypes are respectively calculated, and the cell positive rate is the ratio of the number of positive cells in the visual field to the total number of cells in the visual field;
further specifically, in the step D, positive cells are positioned as two-dimensional cartesian coordinate system data, default data are pixel coordinates based on a rectangular image field of view, positive intercellular distance calculation is performed after the data are converted into length coordinate data according to the magnification, and the following euclidean distance calculation mode is used for calculating the distance:
；
wherein,d（x，y) Representing cellsxAnd cellsyIs used for the distance of euclidean distance,x 1 、x 2 representing cellsxIs defined by the abscissa and the ordinate of (c),y 1 、y 2 representing cellsyAnd the abscissa and ordinate of (c).
Preferably, the immunotherapy in the model for predicting the therapeutic effect of immunotherapy is antibody therapy, and more preferably anti-PD-1 antibody therapy; the disease aimed by the immune treatment effect prediction model is tumor.
In some specific embodiments, the model for predicting the efficacy of an anti-PD 1 antibody drug for treating non-small cell lung cancer is a model for predicting the efficacy of an anti-PD 1 antibody drug for treating non-small cell lung cancer.
Those skilled in the art will appreciate that: the specific embodiment of the invention takes a curative effect prediction model aiming at a first-line immunity combined chemotherapy treatment sample of the advanced stage of the non-small cell lung cancer as an example, but the whole detection method and the model are also suitable for the curative effect prediction of the non-small cell lung cancer perioperative, second-line or post-line treatment to a certain extent.
Those skilled in the art will appreciate that: specific embodiments of the present invention are illustrative of non-small cell lung cancer, but detection methods are applicable to some extent to other tumor types. The tumor may be a benign tumor or a malignant tumor (cancer), including but not limited to a solid tumor or a non-solid tumor.
Those skilled in the art will appreciate that: the technical method used in the invention is mIHC/IF, which is essentially protein detection in situ on tumor tissue, but other methods for protein detection in situ on tissue can also realize detection according to the markers of the invention, and therefore, the detection is also within the protection scope of the invention.
Those skilled in the art will appreciate that: the invention essentially carries out in-situ protein detection on tumor tissues, and protein is derived from translation of mRNA according to a central rule in biology, so that a method for carrying out RNA detection in situ on tissues is combined with the marker for detection, and the method is also within the scope of protection of the invention.
The invention has the beneficial effects that:
compared with the application research and the prior report of the mIHC/IF technology in the NSCLC immunotherapy field, the beneficial effects of the invention are mainly represented by the following three points:
(1) The biomarker is derived from the summary and the summary of a large number of recent literature reports in related fields, and covers a more comprehensive TIME immune cell subset;
(2) The invention fully utilizes the advantage that the mIHC/IF technology can detect space information on the basis of comprehensively analyzing the characteristics of immune cell subsets, establishes a new space relation algorithm, and brings the characteristics of space distribution among immune cells and between tumors and immune cells into the model of the invention;
(3) The invention not only utilizes the machine learning method to perform feature screening and model construction on target protein expression, abundance and spatial distribution features, but also simplifies the model from the clinical application end so as to improve the convenience of the model in use under the real application scene.
In conclusion, the invention utilizes the advantages of mIHC/IF compared with the traditional PD-L1 IHC technology to comprehensively analyze TIME of NSCLC patients, and obtains a more accurate curative effect prediction model aiming at NSCLC immunotherapy.
Drawings
FIG. 1 is a flow and a technical route of the present invention.
FIG. 2 is a panel 1 mIHC staining pattern.
FIG. 3 is a panel 2 mIHC staining chart.
FIG. 4 is a 3 mIHC staining chart of the panel.
FIG. 5 is a panel 4 mIHC staining pattern.
FIG. 6 is a panel 5 mIHC staining pattern.
FIG. 7 is a schematic diagram of tissue splitting.
FIG. 8 is a schematic diagram of cell resolution.
Fig. 9 is a schematic diagram of spatial analysis-radius range count.
FIG. 10 is a schematic of a spatial analysis-nearest neighbor cell.
FIGS. 11-14 show positive rate characteristic effect survival curves.
Fig. 15-16 are spatial feature survival curves.
FIG. 17 is a feature penalty process performed by Lasso Cox.
Fig. 18 shows a model construction weighting coefficient calculation process performed by Lasso Cox.
Figure 19 is a plot of immune profile score, panoscore training set survival.
Figure 20 is a survival curve of the immune profile score Panoscore validation set.
Figure 21 is a total immune profile score Panoscore survival curve.
Fig. 22 shows the effect of differentiation of TPS and corresponding cutoff value cutoff (TPS 1%) on survival curves obtained using the PD-L1 IHC assay.
Fig. 23 shows the effect of differentiation of TPS and corresponding cutoff value cutoff (TPS 50%) on survival curves obtained using the PD-L1 IHC assay.
Figure 24 shows the differentiating effect of immune profile scores obtained in example 2 on PFS survival curves.
Figure 25 shows the differentiating effect of immune profile scores versus PFS survival curve obtained in example 3.
FIG. 26 shows the interstitial region Granzyme B + Differentiation effect of positive rate of cells on PFS survival curve.
FIG. 27 is a 100 μm in Foxp3 + Cell to PANCK + Differentiation effect of shortest distance of cells on PFS survival curve.
FIG. 28 shows the effect of differentiating positive rate of overall CD20+ cells against PFS survival curve.
Detailed Description
The present invention will be described in further detail with reference to the following examples, which are not intended to limit the present invention, but are merely illustrative of the present invention. The experimental methods used in the following examples are not specifically described, but the experimental methods in which specific conditions are not specified in the examples are generally carried out under conventional conditions, and the materials, reagents, etc. used in the following examples are commercially available unless otherwise specified.
The term "panel" as used herein, as known to those skilled in the art, is also referred to as "panel".
The brands of partial reagents adopted in the specific embodiment of the invention are described in Chinese and English forms, and when ambiguity exists, the English brand names are used as the reference.
Example 1
Referring to the technical route shown in fig. 1, the method specifically includes:
1. specimen collection
The experiments used Formalin Fixed Paraffin Embedded (FFPE) puncture samples.
Sample collection criteria: after receiving the samples, the tabletting technicians complete tissue dehydration, wax dipping, embedding and tabletting within 2 working days according to industry specifications (China medical society: clinical technical operation specifications, pathological booklet). The slice is required to be complete, no frame wrinkles exist, and the thickness is 3-5 mu m.
Specimen source: 110 out of 3000+ NSCLC patients were strictly selected. The diagnosis standard refers to the diagnosis and treatment guide 2023 of non-small cell lung cancer of China Society of Clinical Oncology (CSCO).
The initial diagnosis time is from 1 month in 2020 to 12 months in 2021.
Screening criteria: first line immunotherapy based on anti-PD-1 antibody K drug (curida, moesadong) was received.
Advantages and benefits of this step:
the clinical information of the clinical queue is finely screened, so that the baseline background of the clinical queue used by the invention is clear, and a more stable and reliable model is obtained.
2. Multiplex immunofluorescence panel design and staining
2.1 Target selection and panel design
2.1.1 Target selection
Biomarkers in the tumor immune microenvironment of NSCLC patients likely to affect anti-PD-1 treatment:
immune cell subtype markers including CD3 (T cell ), CD4 (Helper T cell ), CD8 (Cytotoxic T cell, cytotoxic T cell), foxp3 (Regulatory T cell ), CD68 (Macrophages, macrophages), CD103 (Tissue Resident Memory T cell ), CD45RO (Memory T cell, memory T cell), CD20 (B cell ), CD138 (Plasma cell), granzyme B/Perforin (Effector T cell ); immune checkpoint/therapeutic targets, including PD-L1, PD-1, LAG-3, TIM-3, TIGIT.
2.1.2 Panel design
Table 1 panel 1
Surface 2 Panel 2
Surface 3 Panel 3
Surface 4 Panel 4
Surface 5 Panel 5
2.2 mIHC staining
The immunocyte subtypes and immunodetection expression of 5 panels were examined in the 110 FFPE specimens described above. The method comprises the following specific steps:
baking slices: the FFPE sample is baked for 20 minutes at 65 ℃;
dewaxing: xylene dewaxing (5 min, 3 replicates);
hydration: gradient ethanol dipsticks (100% 5 min, 95% 5 min, 70% 2 min);
antigen retrieval: immersing the dewaxed and hydrated glass slide in alkaline antigen retrieval liquid (PANOVUE; cat. 0019020500), boiling with high fire in a microwave oven, maintaining with low fire for 15 min, taking out, and naturally cooling to room temperature;
closing: an anti-blocking solution (PANOVUE; cat. 0018001120), shaking at room temperature for 10 min;
incubation resistance: taking PD-1 as an example, PD-1 (CST 43248, 1:100), incubated for 30 minutes at room temperature;
secondary antibody incubation: dripping HRP secondary antibody working solution (PANOVUE; cat. 0013001010), and incubating for 10 minutes at room temperature;
fluorescent staining amplifies the signal: TSA dye PPD520TSA (1:100) was added and incubated for 10 minutes with shaking at room temperature. The 1 XTBST buffer rinse was repeated 3 times.
After each round of antibody staining, antigen retrieval was repeated, and the next round of antibody staining was performed until all markers of a single panel were stained.
Sealing piece: dripping DAPI working solution, and incubating at room temperature; the slides were immersed in 1 XTBST buffer for 3 minutes at room temperature.
Washing the slide with sterilized water for 2 min, dripping super-strong anti-quenching sealing tablet onto the slide with a liquid shifter, and immersing the sample area.
And (5) covering a glass slide and sealing the glass slide. And (5) reading the stained tissue slices, and observing and judging the stained tissue slices under a fluorescence microscope.
Examples of the dyeing result are shown in FIGS. 2 to 6.
Advantages and benefits of this step:
(1) Important biomarkers associated with NSCLC immunotherapy were refined by literature summaries and internal data mining.
(2) Through the strict design of the detection panel, the panel is original, the co-localization of a plurality of biomarkers and the space feature analysis inside the panel are possible, and the biomarkers capable of predicting the curative effect of NSCLC immunotherapy are fully covered.
3. Scanning and reading
3.1 Scanning
The stained FFPE tissue sections were scanned using a high throughput panoramic scanner PanoVIEW VS200 to obtain bright field and fluorescent images of the entire slide.
3.2 Reading sheet
3.2.1 confirmation of image scanning quality
Confirming that focusing is clear, and rescanning a sample which is subjected to scanning blurring and cannot be normally analyzed; confirming that the scan is complete, samples that fail to bring all tissue into scan range require rescanning.
3.2.2 analysis area circling
And presetting an analysis region division rule of the NSCLC puncture sample by a pathologist. After confirming that there is no problem in the image scanning quality, the analysis area is circled. The rule content is as follows:
(1) Areas with no/few DAPI nuclear staining: no DAPI staining, no matter whether other channels are stained or not, all were rejected. Small/sporadic DAPI staining, if blood and muscle tissue, total rejection;
(2) Non-cellular constituent rejection with DAPI staining;
(3) Folding area: removing a folding area caused by tissue flaking partially in the dyeing scanning image;
(4) Removing scattered areas which are free from the whole sample;
(5) Normal epithelial tissue rejection.
Advantages and benefits of this step:
(1) Through scanning and film reading, the image quality is ensured.
(2) Through analysis region division rule formulation and standardization, the whole analysis region can be unified, the comparability between each sample result is stronger, and the result is ensured to be more reliable finally.
4. Data analysis
4.1 tissue type partitioning
The tumor area (tumor tissue) and the interstitial area (tumor interstitial) are subjected to tissue splitting through an artificial neural network-multilayer perceptron algorithm, and corresponding analysis areas are generated for subsequent analysis. The machine learning method used was an artificial neural network-multilayer Perceptron (Artificial Neural Network-Multi-Layer Perceptron, ANN-MLP). Specific references are as follows: m. Jabaruteti and H. Soltanian-Zadeh, 'Medical Image Segmentation Using Artificial Neural Networks', artificial Neural Networks-Methodological Advances and Biomedical applications. InTech, apr.11, 2011. Doi: 10.5772/16103.
Tumor tissue: areas stained with DAPI and PANCK (one of the markers described above) have a larger cell volume and a disordered tissue dispersion.
Tumor stroma: areas with DAPI staining and no PANCK staining, with smaller cell volumes and clean tissue.
The effect after splitting is shown in fig. 7, the right graph shows DAPI (fluorescence channel a) and PANCK (fluorescence channel B) fluorescence channels, the left graph shows that the region type a is the identified tumor region, the region type B is the interstitial region, and the region type C is the non-tissue region.
4.2 Cell resolution
Designating a core channel: typically the DAPI channel, is used to identify the nucleus.
Setting a threshold value: according to the expression intensity of DAPI, a threshold value is set, so that the area with low DAPI expression is prevented from being identified as the nucleus.
Degree of resolution: the watershed algorithm is used for accurately splitting cells, the effect after splitting is shown in figure 8, the cell splitting type A is a cell nucleus, the cell splitting type B is a split cell, each cell comprises two circles, the inner circle is a cell nucleus area, and the outer circle is a cell cytoplasm area. Specific algorithm references: ali S, madabhurhi A. An integrated region-, boundary-, shape-based active contour for multiple object overlap resolution in histological image IEEE Trans Med imaging 2012 Jul, 31 (7): 1448-60. Doi: 10.1109/TMI.2012.2190089. Epub 2012 Apr 5. PMID: 22498689.
4.3 Feature calculation and result output
4.3.1 And (5) calculating the corresponding cell positive rate by respectively measuring the characteristics of fluorophores in different areas of the tumor.
Cell positive rate= (Σc_p)/(Σc)
Wherein, the cells refer to cells of different subtypes (the proportion of the different subtypes is calculated respectively), and C is the cell count under the visual field; c_p is the phenotype cell count positive for staining in C.
Single target: CD3, CD4, CD8, foxp3, CD68, CD103, CD45RO, CD20, CD138, granzyme B, perforin, PD-L1, PD-1, LAG-3, TIM-3, TIGIT.
Target co-localization: CD8 + PD-L1 + ，CD68 + PD-L1 + ，CD8 + PD-1 + ，CD8 + CD103 + ，CD8 + Perforin + ，CD8 + Granzyme B + ，CD4 + Foxp3 + ，CD8 + Foxp3 + ，CD8 + CD45RO + ，CD8 + LAG-3 + ，CD8 + TIM-3 + ，CD8 + TIGIT + Etc.
The distinguishing effect of some biomarkers on the clinical survival data of the sample is shown in the following figures:
positive rate characteristic distinguishing effect is exemplified (as in fig. 11 to 14): integral PD-L1 + PANCK + Cell positive rate, overall CD45RO + Cell positive rate, interstitial region CD8 + Cell positive rate and tumor zone CD68 + PD-L1 + Cell positive rate, etc. can distinguish patient PFS.
4.3.2 The following characteristics of the spatial tissue distribution of cells were analyzed using the Inform software.
The cell-to-cell distance data are two-dimensional Cartesian coordinate system data, default data are pixel coordinates based on a rectangular image visual field, and the data can be converted into corresponding length coordinate data according to the magnification ratio and then calculated. Simple illustrations are shown in fig. 9-10.
The distance is calculated using the following Euclidean distance calculation method:
；
wherein,d（x，y) Representing cellsxAnd cellsyIs used for the distance of euclidean distance,x 1 、x 2 representing cellsxIs defined by the abscissa and the ordinate of (c),y 1 、y 2 representing cellsyAnd the abscissa and ordinate of (c).
Can be determined by itself by the distribution of cells in the field of view.
(1) By CD8 + Cell-centered, separate calculation of CD8 + Cell to PANCK + Shortest distance to tumor and to CD68 + Shortest distance of cells (pairing of two values, subsequent calculation of ratio).
(2) With PD-L1 + PD-1 in the radius 5/10/15/20 μm range with the cells as the center + Cell number and cell number to PD-L1 + Distance of cells.
(3) By CD8 + PD-L1 with cell as center and radius of 30 μm + And CD68 + Cell number and CD8 + Distance of cells.
(4) In CD4 + Cell-centered, separate calculation of CD4 + To PANCK + Shortest distance to tumor and to Foxp3 + Shortest distance of cells (pairing of two values, subsequent calculation of ratio).
(5) By CD8 + Cell-centered, separate calculation of CD8 + And PANCK + Shortest distance to tumor and to Foxp3 + Shortest distance of cells (pairing of two values, subsequent calculation of ratio).
Spatial feature discrimination effect example (as in fig. 15-16): CD8 + P in the 30 μm range of cellsD-L1 + Cell number and CD4 + To PANCK + Shortest distance sum to Foxp3 of tumor cells + The ratio of shortest distances of cells, etc. can distinguish patient PFS.
Advantages and benefits of this step:
(1) By utilizing the application of artificial intelligence and machine interpretation in the analysis flow, the stability of the analysis result is ensured, and the subjectivity of human interpretation is avoided.
(2) Through the co-localization of a plurality of biomarkers and the spatial feature analysis, the more abundant and more accurate biomarkers related to the curative effect of NSCLC immunotherapy are obtained, which is information that cannot be obtained by the traditional IHC method.
5. Construction of immune treatment efficacy prediction model
The invention uses R language, and utilizes the biomarker or the positive rate of the cell type, the spatial characteristics and the clinical survival information progression-free survival time (PFS) corresponding to the sample to explore the correlation between the immune characteristic model and the clinical index. And calculating weights for each index through a machine learning algorithm, and calculating immune characteristic scores (Panoscore) through a weighted summation mode. The method comprises the following steps:
5.1 Sample quality control
And (3) performing quality control on the detected sample according to the following analysis results:
(1) Whether the total number of cells in the slice is greater than 1000.
(2) Whether the tumor cell ratio in the slice is more than 10 percent.
(3) Whether the cell phenotype interpretation results in a feature number of 0 is less than 20% of the total feature number.
And if the sample quality control result does not meet any of the above-mentioned items, namely the samples are considered to have problems, the subsequent analysis is eliminated, and finally 102 samples remain.
5.2 Data preprocessing
The following pre-treatments were performed on the results of the mIHC/IF image analysis: missing value processing, outlier processing, feature conversion, normalization processing, and the like.
5.3 Feature screening
The feature screening mainly comprises 2 steps: (1) Based on survival data information in clinical information, classifying the samples into 2 groups, and screening the characteristics with obvious differences among the 2 groups according to statistical assumptions; (2) And further screening the features according to the inherent correlation of the screened features.
5.4 Model construction
(1) 102 samples are combined according to a training set: the validation set is 2:1, and randomly grouping. Training set 68 cases, validation set 34 cases.
(2) And punishment and screening are carried out on the characteristics by using a Lasso Cox method in a training set, and finally, a NSCLC immunotherapy efficacy prediction model based on sample characteristics and PFS survival information is constructed. The feature penalty process and model building process (i.e., the weighting factor calculation process) are shown in fig. 17-18. The effect of the training of the obtained model on the survival data of the training set samples is shown in fig. 19.
(3) The modeling screening model was repeated n times for k-fold to obtain the following immune scoring model: panoscore=0.42× (interstitial region Granzyme B + Positive rate of cells) +0.63× (tumor zone CD 68) + PD-L1 + Positive rate of cells) +0.20× (interstitial region CD8 + Cell positive) +0.10× (global CD45 RO) + Positive rate of cells) +0.10× (Whole area PANCK + PD-L1 + Cell positive) +0.81× (Foxp 3 within 100 microns) + Cell to PANCK + Shortest distance of cells).
(4) And verifying the model in a verification set, and judging whether the model can be stabilized in the verification set.
The model set obtained in the above step was used in the validation set and the same cut-off value cutoff was used to differentiate patients, and the survival curve obtained was as shown in fig. 20. The survival curves of the training set and the set of validation sets (i.e., the total set) at the cutoff value cutoff described above are shown in fig. 21.
Advantages and benefits of this step:
(1) The influence of the abnormal detection result of the real-world abnormal sample on the model is reduced through quality control and data preprocessing of the sample analysis result, so that the model has better universality.
(2) Through multiple rounds of feature screening, biomarkers truly and closely related to the treatment effect are screened out as much as possible; the screened characteristics are put into the model for modeling, so that the obtained model is more robust and is fit with the real situation
(3) The random components of the sample are used as a training set verification set, so that a model trained in the training set is verified, and the model is stable as much as possible.
102 samples incorporating the invention were taken to differentiate effects according to survival curves provided by clinically existing biomarkers (i.e. TPS scores obtained from PD-L1 IHC assays) and corresponding cut-off values cutoff (1% versus 50%) as shown in figures 22-23.
To sum up: the PFS distinguishing effect of the Panoscore and TPS on the total set is obvious, and the distinguishing effect of the constructed Panoscore model immune score is obviously better than the TPS score used clinically at present.
Example 2
The difference from example 1 is that: interstitial region Granzyme B + The coefficient of the positive rate of the cells selects the minimum value of a feasible coefficient interval; tumor zone CD68 + PD-L1 + The coefficient of the positive rate of the cells selects the maximum value of the feasible coefficient interval.
Model of immune score: panoscore=0.20× (interstitial region Granzyme B + Positive rate of cells) +0.80× (tumor zone CD 68) + PD-L1 + Positive rate of cells) +0.20× (interstitial region CD8 + Cell positive) +0.10× (global CD45 RO) + Positive rate of cells) +0.10× (Whole area PANCK + PD-L1 + Cell positive) +0.81× (Foxp 3 within 100 microns) + Cell to PANCK + Shortest distance of cells).
The results are shown in FIG. 24. The differentiation effect of the immune scores of the Panoscore model constructed by the invention is obviously better than the TPS scores used clinically at present.
Example 3
The difference from example 1 is that: interstitial region Granzyme B + The coefficient of the positive rate of the cells selects the maximum value of a feasible coefficient interval; tumor zone CD68 + PD-L1 + Cell yangThe coefficient of the sex ratio selects the minimum value of the feasible coefficient interval.
Model of immune score: panoscore=0.45× (interstitial region Granzyme B + Positive rate of cells) +0.40× (tumor zone CD 68) + PD-L1 + Positive rate of cells) +0.30× (interstitial region CD8 + Cell positive) +0.10× (global CD45 RO) + Positive rate of cells) +0.10× (Whole area PANCK + PD-L1 + Cell positive) +0.81× (Foxp 3 within 100 microns) + Cell to PANCK + Shortest distance of cells).
The results are shown in FIG. 25. The differentiation effect of the immune scores of the Panoscore model constructed by the invention is obviously better than the TPS scores used clinically at present.
Comparative examples were set as follows with reference to example 1:
comparative example 1
Using only the interstitial region Granzyme B + The effect of differentiation of positive rate of cells against PFS survival curve is shown in fig. 26.
Comparative example 2
Using only the interstitial region CD8 + The effect of differentiating positive rate of cells against PFS survival curve is shown in FIG. 14.
Comparative example 3
Tumor zone CD68 + PD-L1 + The effect of differentiating positive rate of cells against PFS survival curve is shown in FIG. 12.
Comparative example 4
Integral CD45RO + The effect of differentiating positive rate of cells against PFS survival curve is shown in FIG. 11.
Comparative example 5
Integral PANCK + PD-L1 + The effect of differentiation of positive rate of cells against the survival curve of PFS is shown in fig. 13.
Comparative example 6
Foxp3 within 100 microns + Cell to PANCK + The effect of the shortest distance of cells on the differentiation of the survival curve of PFS is shown in figure 27.
Comparative example 7
Integral CD20 + The effect of differentiating positive rate of cells against PFS survival curve is shown in FIG. 28.
The model constructed by the markers together obviously has better distinguishing effect and better clinical application prospect.
The above examples are provided for illustrating the present invention and are not intended to limit the present invention, and those skilled in the art may substitute or optimize the conventional means according to the description of the present invention, and the obtained technical solutions are also within the scope of the present invention.
Claims (14)
1. A marker combination for predicting the efficacy of an immunotherapy, comprising the following markers: granzyme B, CD68, CD8, PD-L1, PANCK, foxp3 and CD45RO.
2. Use of the marker combination of claim 1 for constructing an immunotherapeutic efficacy prediction model.
3. An immunotherapeutic efficacy prediction model constructed by the marker combination of claim 1.
4. The model for predicting therapeutic efficacy of immunotherapy according to claim 3, wherein the model is an integral PD-L1 model + PANCK + Cell positive rate, overall CD45RO + Cell positive rate, interstitial region CD8 + Cell positive rate, interstitial region Granzyme B + Cell positive rate, tumor zone CD68 + PD-L1 + Cell positive rate and Foxp3 within 100 microns + Cell to PANCK + The shortest distance of cells was constructed.
5. The model of claim 4, wherein the model is used to calculate an immune score by: (0.20-0.45) × (interstitial region Granzyme B + Positive rate of cells) + (0.40-0.80) × (tumor area CD 68) + PD-L1 + Positive rate of cells) + (0.10-0.20) × (interstitial region CD 8) + Cell positive) + (0.05-0.15) × (Whole CD45 RO) + Positive rate of cells) + (0.05-0.15) x (integral region PANCK + PD-L1 + Cell positive) + (0.70-0.90) × (Foxp 3 within 100 microns) + Cell to PANCK + Shortest distance of cells).
6. The model of claim 5, wherein the predictive cut-off is 0.5.
7. The model for predicting the efficacy of an immunotherapy according to claim 6, wherein the immune score is calculated by: 0.42× (interstitial region Granzyme B + Positive rate of cells) +0.63× (tumor zone CD 68) + PD-L1 + Positive rate of cells) +0.20× (interstitial region CD8 + Cell positive) +0.10× (global CD45 RO) + Positive rate of cells) +0.10× (Whole area PANCK + PD-L1 + Cell positive) +0.81× (Foxp 3 within 100 microns) + Cell to PANCK + Shortest distance of cells).
8. The model for predicting the efficacy of an immunotherapy according to any one of claims 3 to 7, which is a model for predicting the efficacy of an anti-PD-1 antibody drug therapy.
9. The model for predicting the efficacy of an immunotherapy according to claim 8, wherein the model predicts the efficacy of an anti-PD-1 antibody drug for treating non-small cell lung cancer.
10. The method for constructing an immunotherapy efficacy prediction model according to any one of claims 3 to 9, comprising:
step (1): FFPE specimen mIHC is dyed;
step (2): scanning and film reading;
step (3): analyzing data;
step (4): and (5) calculating an immune score.
11. The method of claim 10, wherein the step (1) comprises: baking slices; dewaxing; hydrating; antigen retrieval; closing; incubation with primary antibody; incubating the secondary antibody; amplifying the signal by fluorescent staining; repeating antigen repair after each round of antibody staining, and carrying out the next round of antibody staining; sealing; and reading the film.
12. The method of claim 11, wherein the step (2) includes: after scanning, clearly and completely dividing analysis areas of the images, wherein the dividing rules are as follows:
1) Areas with no or little DAPI nuclear staining: no DAPI staining is carried out, no matter whether other channels are stained or not, all the channels are removed; small or sporadic DAPI staining, if blood and muscle tissue, total rejection;
2) Non-cellular constituent rejection with DAPI staining;
3) Folding area: removing a folding area caused by tissue flaking partially in the dyeing scanning image;
4) Removing scattered areas which are free from the whole sample;
5) Normal epithelial tissue rejection.
13. The method of claim 12, wherein the step (3) includes:
A. tissue splitting: tumor tissue and tumor stroma resolution;
B. cell resolution: cell nucleus and cytoplasm resolution;
C. calculating the characteristics of the cell positive rate;
D. positive cell positioning, calculating the distance between positive cells, and analyzing the spatial tissue distribution characteristics of cells.
14. The method of claim 13, wherein in a, the splitting is performed by DAPI staining and PANCK staining;
in the step B, splitting is carried out through the expression intensity of DAPI;
in the step C, positive cells of different subtypes are respectively calculated, and the cell positive rate is the ratio of the number of positive cells in the visual field to the total number of cells in the visual field;
in the step D, positive cells are positioned as two-dimensional Cartesian coordinate system data, default data are pixel coordinates based on a rectangular image visual field, positive cell distance calculation is carried out after the data are converted into length coordinate data according to the magnification, and the following Euclidean distance calculation mode is used for calculating the distance:
；
wherein,d（x，y) Representing cellsxAnd cellsyIs used for the distance of euclidean distance,x 1 、x 2 representing cellsxIs defined by the abscissa and the ordinate of (c),y 1 、y 2 representing cellsyAnd the abscissa and ordinate of (c).
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410182047.8A CN117741148B (en) | 2024-02-19 | 2024-02-19 | Marker combination for predicting curative effect of immunotherapy and model construction method and application |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410182047.8A CN117741148B (en) | 2024-02-19 | 2024-02-19 | Marker combination for predicting curative effect of immunotherapy and model construction method and application |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117741148A true CN117741148A (en) | 2024-03-22 |
| CN117741148B CN117741148B (en) | 2024-06-14 |
Family
ID=90279817
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410182047.8A Active CN117741148B (en) | 2024-02-19 | 2024-02-19 | Marker combination for predicting curative effect of immunotherapy and model construction method and application |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117741148B (en) |
Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105209919A (en) * | 2013-03-15 | 2015-12-30 | 豪夫迈·罗氏有限公司 | Biomarkers and methods of treating PD-1 and PD-L1 related conditions |
| CN108251523A (en) * | 2016-12-27 | 2018-07-06 | 山东省医学科学院基础医学研究所 | A kind of non-small cell lung cancer molecular marker and its application |
| CN109596831A (en) * | 2019-01-14 | 2019-04-09 | 臻悦生物科技江苏有限公司 | The Multiple immunizations histochemical analysis kit and its application method of a kind of lung cancer and application |
| CN111638357A (en) * | 2020-07-01 | 2020-09-08 | 山东凯歌智能机器有限公司 | Immunofluorescence kit and method for E-Cadherin mutation of peripheral blood circulating tumor cells of patient with non-small cell lung cancer |
| CN113192560A (en) * | 2021-03-02 | 2021-07-30 | 郑州大学第一附属医院 | Construction method of hepatocellular carcinoma typing system based on iron death process |
| CN114509567A (en) * | 2022-02-28 | 2022-05-17 | 成都福实生物科技有限公司 | Application of reagent for detecting spatial distribution of immune check points in preparation of tumor treatment and prognosis diagnosis products |
| CN115902215A (en) * | 2022-11-29 | 2023-04-04 | 南京医科大学 | Non-small cell lung cancer prognosis marker and application thereof |
| CN116206681A (en) * | 2021-11-30 | 2023-06-02 | 中南大学湘雅医院 | Method for evaluating prognostic gene pair value of immune infiltration cell model |
| CN116265947A (en) * | 2021-12-17 | 2023-06-20 | 许大康 | Immune biomarker combination based on CD8/FOXP3 and PD-1 and PD-L1 and application thereof |
| CN116466089A (en) * | 2023-04-27 | 2023-07-21 | 杭州华得森生物技术有限公司 | SCLC molecular typing-based CTC detection kit adopting microfluidic chip and multiple immunofluorescence probe technology and identification method thereof |
| CN116479126A (en) * | 2023-03-29 | 2023-07-25 | 首都医科大学附属北京胸科医院 | Marker for predicting curative effect of lung cancer immunotherapy and construction method of scoring system |
| CN116819066A (en) * | 2023-06-30 | 2023-09-29 | 复旦大学附属中山医院 | Liver cancer microenvironment detection kit |
-
2024
- 2024-02-19 CN CN202410182047.8A patent/CN117741148B/en active Active
Patent Citations (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20160222118A1 (en) * | 2013-03-15 | 2016-08-04 | Genentech, Inc. | Biomarkers and methods of treating pd-1 and pd-l1 related conditions |
| CN105209919A (en) * | 2013-03-15 | 2015-12-30 | 豪夫迈·罗氏有限公司 | Biomarkers and methods of treating PD-1 and PD-L1 related conditions |
| CN108251523A (en) * | 2016-12-27 | 2018-07-06 | 山东省医学科学院基础医学研究所 | A kind of non-small cell lung cancer molecular marker and its application |
| CN109596831A (en) * | 2019-01-14 | 2019-04-09 | 臻悦生物科技江苏有限公司 | The Multiple immunizations histochemical analysis kit and its application method of a kind of lung cancer and application |
| CN111638357A (en) * | 2020-07-01 | 2020-09-08 | 山东凯歌智能机器有限公司 | Immunofluorescence kit and method for E-Cadherin mutation of peripheral blood circulating tumor cells of patient with non-small cell lung cancer |
| CN113192560A (en) * | 2021-03-02 | 2021-07-30 | 郑州大学第一附属医院 | Construction method of hepatocellular carcinoma typing system based on iron death process |
| CN116206681A (en) * | 2021-11-30 | 2023-06-02 | 中南大学湘雅医院 | Method for evaluating prognostic gene pair value of immune infiltration cell model |
| CN116265947A (en) * | 2021-12-17 | 2023-06-20 | 许大康 | Immune biomarker combination based on CD8/FOXP3 and PD-1 and PD-L1 and application thereof |
| CN114509567A (en) * | 2022-02-28 | 2022-05-17 | 成都福实生物科技有限公司 | Application of reagent for detecting spatial distribution of immune check points in preparation of tumor treatment and prognosis diagnosis products |
| CN115902215A (en) * | 2022-11-29 | 2023-04-04 | 南京医科大学 | Non-small cell lung cancer prognosis marker and application thereof |
| CN116479126A (en) * | 2023-03-29 | 2023-07-25 | 首都医科大学附属北京胸科医院 | Marker for predicting curative effect of lung cancer immunotherapy and construction method of scoring system |
| CN116466089A (en) * | 2023-04-27 | 2023-07-21 | 杭州华得森生物技术有限公司 | SCLC molecular typing-based CTC detection kit adopting microfluidic chip and multiple immunofluorescence probe technology and identification method thereof |
| CN116819066A (en) * | 2023-06-30 | 2023-09-29 | 复旦大学附属中山医院 | Liver cancer microenvironment detection kit |
Non-Patent Citations (3)
| Title |
|---|
| EDWIN ROGER PARRA ET AL: "Immuno-profiling and cellular spatial analysis using five immune oncology multiplex immunofluorescence panels for paraffin tumor tissue", SCIENTIFIC REPORTS, vol. 11, no. 1, 19 April 2021 (2021-04-19), pages 8511 * |
| PARRA, ER ET AL: "Effect of neoadjuvant chemotherapy on the immune microenvironment in non-small cell lung carcinomas as determined by multiplex immunofluorescence and image analysis approaches", JOURNAL FOR IMMUNOTHERAPY OF CANCER, vol. 6, 6 June 2018 (2018-06-06), pages 48 * |
| 尹婕;何志昂;张毅;王丽萍;: "非小细胞肺癌组织中Foxp3、PD-1和CD45RO蛋白的表达", 郑州大学学报(医学版), no. 02, 30 April 2020 (2020-04-30), pages 196 - 200 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117741148B (en) | 2024-06-14 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| AbdulJabbar et al. | Geospatial immune variability illuminates differential evolution of lung adenocarcinoma | |
| US20220051804A1 (en) | Image Analysis for Breast Cancer Prognosis | |
| Lloyd et al. | Using image analysis as a tool for assessment of prognostic and predictive biomarkers for breast cancer: How reliable is it? | |
| EP2861986B1 (en) | Method for quantifying immune cells in tumoral tissues and its applications | |
| US20110111435A1 (en) | Detecting Cell Surface Markers | |
| US11721427B2 (en) | Computational pathology approach for retrospective analysis of tissue-based companion diagnostic driven clinical trial studies | |
| AU2015357091A1 (en) | Systems and methods for early-stage cancer prognosis | |
| Holten‐Rossing et al. | Application of automated image analysis reduces the workload of manual screening of sentinel lymph node biopsies in breast cancer | |
| Ali et al. | Astronomical algorithms for automated analysis of tissue protein expression in breast cancer | |
| CN113933505B (en) | TIIC index for predicting curative effect of gastric cancer immunotherapy by multidimensional analysis and application thereof | |
| Yaseen et al. | Validation of an accurate automated multiplex immunofluorescence method for immuno-profiling melanoma | |
| Qian et al. | Computerized assessment of the tumor-stromal ratio and proposal of a novel nomogram for predicting survival in invasive breast cancer | |
| US20210249101A1 (en) | Systems and methods for predictive molecular biomarker identification and quantification from morphology changes in histopathology tissue | |
| Xie et al. | Artificial intelligence scale-invariant feature transform algorithm-based system to improve the calculation accuracy of Ki-67 index in invasive breast cancer: a multicenter retrospective study | |
| Roberts et al. | Reliability of a computational platform as a surrogate for manually interpreted immunohistochemical markers in breast tumor tissue microarrays | |
| Yaghjyan et al. | Reliability of CD44, CD24, and ALDH1A1 immunohistochemical staining: Pathologist assessment compared to quantitative image analysis | |
| CN117741148B (en) | Marker combination for predicting curative effect of immunotherapy and model construction method and application | |
| Narayanan et al. | Unmasking the tissue microecology of ductal carcinoma in situ with deep learning | |
| Coakley et al. | A comparison of imaging software and conventional cell counting in determining melanocyte density in photodamaged control sample and melanoma in situ biopsies | |
| Fernandez et al. | Analytical validation of the PreciseDx digital prognostic breast cancer test in early-stage breast cancer | |
| CN113496769B (en) | Triple comprehensive tumor analysis system based on pathological tissue and application | |
| Sobral-Leite et al. | Artificial intelligence-based morphometric signature to identify ductal carcinoma in situ with low risk of progression to invasive breast cancer | |
| CN113496768A (en) | Bigeminal comprehensive tumor analysis system based on pathological tissues and application | |
| Su et al. | Multiplex immunohistochemistry and high-throughput image analysis for evaluation of spatial tumor immune cell markers in human breast cancer | |
| RU2748716C1 (en) | Breast cancer prognosis method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |