Marker combination for predicting curative effect of melanoma anti-PD-1 immunotherapy and application of marker combination

Abstract

The invention relates to a marker combination for predicting curative effect of melanoma anti-PD-1 immunotherapy and application of the marker combination. Specifically, more abundant and more accurate biomarkers (S100, CD68, CD3, CD20, PD-L1 and PD-1) related to the curative effect of the melanoma immunotherapy are obtained through co-localization of a plurality of biomarkers and spatial feature analysis, and a prediction model for evaluating the curative effect of the melanoma anti-PD-1 immunotherapy is prepared by using the biomarkers. The model has good sensitivity and specificity, and has a wide application prospect.

Description

Technical Field

[0001] The present invention relates to the field of biological detection, and specifically, to a marker combination for predicting the efficacy of anti-PD-1 immunotherapy for melanoma and its application. Background Art

[0002] Melanoma is a highly malignant tumor originating from melanocytes. It can occur in tissues and organs such as the skin, extremities, mucous membranes, and uveal tract. It has a high degree of malignancy and a poor prognosis.

[0003] According to the site of onset of melanoma and the degree of sun damage, melanoma is divided into four different subtypes: mucosal, acral, chronic sun damage and non-chronic sun damage. The distribution of each subtype has obvious racial differences. Acral melanoma is a high-incidence subtype in the Asian population, and it is prone to occur on the palms, soles and subungual areas. Mucosal melanoma is prone to occur in hidden parts such as the oral cavity, nasal cavity, vagina, esophagus, and intestines. It is a rare subtype in the European and American population, accounting for only 1%-2% of all melanomas. It is the second most common melanoma subtype in the Asian population, accounting for about 20%-30%. Due to its anatomical location, it is easy to delay medical treatment and is usually diagnosed in the late stage of the disease, with a poor prognosis. The 5-year overall survival rate is only about 25%, while the 5-year overall survival rate of cutaneous melanoma is 54% to 80%.

[0004] So far, the evidence-based medicine used in the international melanoma diagnosis and treatment guidelines is mainly based on the research of cutaneous melanoma. The treatment options provided by the guidelines cannot meet the needs of acral and mucosal melanoma, and most advanced patients face the dilemma of "no cure". For example, about 50% of patients with cutaneous melanoma carry BRAF gene mutations. For these patients, BRAF inhibitors combined with MEK inhibitors have significant therapeutic effects. However, in acral melanoma, the frequency of BRAF gene mutations is only 15.5%; in mucosal melanoma, the frequency of BRAF gene mutations is only 12.5%, and a certain proportion of acral and mucosal patients lack sensitivity to BRAF inhibitors due to special mutations.

[0005] In recent years, immune checkpoint inhibitors represented by PD-1 have achieved great success in the treatment of melanoma. The results of several large-scale phase III clinical studies have shown that the first-line treatment represented by PD-1 monoclonal antibodies has an effective rate of up to 26%-44%, which is 4 times higher than that of chemotherapy drugs, and the five-year survival rate is as high as 34-42%. Two clinical studies of anti-PD-1 monoclonal antibodies conducted in my country (KEYNOTE-151 and POLARIS-01) showed that the effective rate of acral melanoma was 15.8-23.1%, and the effective rate of mucosal melanoma was only 0-13.3%. The above data show that only a small number of patients with acral and mucosal melanoma can benefit from the current immunotherapy. This is not surprising considering the complex process of anti-tumor immune response, which is not only related to tumor cells, but also to many factors related to the tumor microenvironment and the entire body.

[0006] Given the significant differences in melanoma subtypes between Asian and European populations, as well as their different ICI response rates, the main challenge currently facing this field is to identify and develop biomarkers that can predict the efficacy of ICI in Asian melanoma, so as to accurately identify tumor patients who will benefit from immunotherapy. Summary of the invention

[0007] In order to fill the gap in the prior art, the present invention provides the following technical solutions:

[0008] The first aspect of the present invention provides a use of a detection reagent in the preparation of a kit for predicting the efficacy of anti-PD-1 immunotherapy for melanoma, wherein the detection reagent is an immunofluorescence reagent, and the biomarkers are CD68, CD3, S100, CD20, PDL1 and PD1.

[0009] In one embodiment, the application includes using the immunofluorescence reagent to stain a patient's melanoma sample, and predicting the efficacy of the patient's anti-PD-1 immunotherapy by identifying and quantitatively calculating the number and percentage of CD3, CD20, CD68, PD-L1, PD-1, PD-L1, PD-1+CD68, PD-1+CD3, PD-1+CD20, PD-L1+CD68, PD-L1+CD3, and PD-L1+CD20 positive cells in the tumor tissue. The calculation includes calculating the immunotherapy prediction score:

[0010] Immunotherapy efficacy score = number of biomarker-positive cells / total number of DAPI-stained (S100, CD68, CD3, CD20, PD-L1, and PD-1)-positive cells;

[0011] or,

[0012] Immunotherapy efficacy score = number of biomarker-positive cells (cells within 0-25 um around S100+ cells) / total number of DAPI-stained (S100, CD68, CD3, CD20, PD-L1 and PD-1)-positive cells (cells within 0-25 um around S100+ cells);

[0013] or,

[0014] Immunotherapy efficacy score = number of biomarker-positive cells (cells within 25-50 um around S100+ cells) / total number of DAPI-stained (S100, CD68, CD3, CD20, PD-L1 and PD-1)-positive cells (cells within 25-50 um around S100+ cells);

[0015] The efficacy of immunotherapy in patients is evaluated based on the Cutoff value (predicted critical value) of the immunotherapy efficacy score corresponding to different biomarkers. Patients with values ​​higher than the cutoff value are judged as those who benefit from immunotherapy, and patients with values ​​lower than the cutoff value are judged as those who do not benefit from immunotherapy (ineffective).

[0016] In a preferred embodiment, the immunotherapy efficacy score = PD1+CD3+ cell number (cells within 25-50um around S100+ cells) / DAPI-labeled cell number (cells within 25-50um around S100+ cells), and the cutoff value is 0.05.

[0017] The second aspect of the present invention provides a melanoma anti-PD-1 immunotherapy efficacy prediction model, wherein the model is: immunotherapy efficacy score = number of biomarker-positive cells / total number of DAPI-stained (S100, CD68, CD3, CD20, PD-L1 and PD-1) positive cells;

[0018] or,

[0019] Immunotherapy efficacy score = number of biomarker-positive cells (cells within 0-25 um around S100+ cells) / total number of DAPI-stained (S100, CD68, CD3, CD20, PD-L1 and PD-1)-positive cells (cells within 0-25 um around S100+ cells);

[0020] or,

[0021] Immunotherapy efficacy score = number of biomarker-positive cells (cells within 25-50 um around S100+ cells) / total number of DAPI-stained (S100, CD68, CD3, CD20, PD-L1 and PD-1)-positive cells (cells within 25-50 um around S100+ cells);

[0022] The model is used as follows: the immunotherapy efficacy of patients is evaluated according to the Cutoff value (prediction critical value) of the immunotherapy efficacy score corresponding to different biomarkers. Patients with values ​​higher than the cutoff value are judged as the group that benefits from immunotherapy, and patients with values ​​lower than the cutoff value are judged as the group that does not benefit from immunotherapy (ineffective).

[0023] In a preferred embodiment, the immunotherapy efficacy score = PD1+CD3+ cell number (cells within 25-50um around S100+ cells) / DAPI-labeled cell number (cells within 25-50um around S100+ cells), and the cutoff value is 0.05.

[0024] The third aspect of the present invention provides a kit for predicting the efficacy of anti-PD-1 immunotherapy for melanoma, the kit comprising a detection reagent, the detection reagent is an immunofluorescence reagent, and the biomarkers are CD68, CD3, S100, CD20, PDL1 and PD1.

[0025] In one embodiment, the kit is used as follows: the immunofluorescence reagent is used to stain the patient's melanoma sample, and the number and percentage of CD3, CD20, CD68, PD-L1, PD-1, PD-L1, PD-1+CD68, PD-1+CD3, PD-1+CD20, PD-L1+CD68, PD-L1+CD3, and PD-L1+CD20 positive cells in the tumor tissue are identified and quantitatively calculated to predict the efficacy of anti-PD-1 immunotherapy for the patient.

[0026] Compared with the prior art, the present invention has the following significant improvements:

[0027] (1) The biomarkers involved in the present invention are derived from the summary and compilation of a large number of recent literature reports in related fields, and the TIME immune cell subsets it covers are more comprehensive;

[0028] (2) Based on the comprehensive analysis of immune cell subset characteristics, the present invention also fully utilizes the advantage of mIHC / IF technology in detecting spatial information, establishes a new spatial relationship algorithm, and incorporates the spatial distribution characteristics between immune cells and between tumors and immune cells into the model of the present invention;

[0029] (3) The present invention not only uses machine learning methods to perform feature screening and model construction on target protein expression, abundance and spatial distribution characteristics, but also simplifies the model from the perspective of clinical application to improve the convenience of using the model in real application scenarios;

[0030] (4) By using artificial intelligence and machine interpretation in the analysis process, the stability of the analysis results is ensured, avoiding the subjectivity of human interpretation;

[0031] (5) Through the co-localization of multiple biomarkers and spatial feature analysis, richer and more accurate biomarkers related to the efficacy of melanoma immunotherapy were obtained, which is information that cannot be obtained by traditional IHC methods. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0033] Figure 1 .Analysis of immune cell populations in melanoma patients with partial response (PR) and progressive disease (PD) after anti-PD-1 treatment is shown. (A) Representative mIF (multiplex immunofluorescence) images of FFPE melanoma tissue samples from patients who showed PR and PD to anti-PD-1 treatment are shown. (B) Total cell numbers in PR and PD samples were quantified. (C) Various immune cell populations and their subsets were quantified as percentages of total cells in PR and PD samples;

[0034] Figure 2 .Analysis of the spatial proximity of immune cells to melanoma cells in patients treated with anti-PD-1 therapy. (A) Representative mIF images showing the spatial distribution of immune cells around S100+ melanoma cells. (B) Quantification of PD-1+ and PD-1+CD3+ cells within 25 μm around S100+ melanoma cells in PR and PD patients. (C) Spatial analysis of PD-1+, PD-1+CD3+, and PD-1+CD20+ cells within 25–50 μm around S100+ melanoma cells.

[0035] Figure 3. Predictive value of tumor-infiltrating immune cells for clinical benefit of anti-PD-1 therapy. (A) ROC curves show the predictive ability of different biomarkers for PFS (progression-free survival), including PD-1, PD-1+CD3, PD-1_25, PD-1_50, PD-1+CD3_25, PD-1+CD3_50, and PD-1+CD20_50 immune cells. (BH) Kaplan-Meier survival analysis of PFS stratified by the proportion of immune cells within a specific distance from melanoma cells: PD-1 (B), PD-1+CD3 (C), PD-1_25 (D), PD-1_50 (E), PD-1+CD3_25 (F), PD-1+CD3_50 (G). ROC, receiver operating characteristic curve; AUC, area under the curve; PFS, progression-free survival. DETAILED DESCRIPTION

[0036] The preferred embodiments of the present invention are described below in conjunction with the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.

[0037] Example 1 Screening of biomarkers for efficacy of melanoma immunotherapy

[0038] Immunotherapy for melanoma involves the study of the relationship between the immune microenvironment and tumor cells. For melanoma cells, the biomarker S100 is usually used to mark the tumor, and for the microenvironment, there are many types of immune and supporting cells. Among the immune cells, there are T cells, B cells, dendritic cells, macrophages, NK cells, granulocytes, etc., and the supporting cells related to immunity involve immune-associated fibroblasts (CAF). The biomarkers for these cells include CD3, CD4, CD8, FOXP3, etc. The tumor-associated tertiary lymph nodes composed of B cells and dendritic cells contain CD20 CD21, CD23, macrophages include CD68, CD163, HLA-DR, etc., and NK cells and granulocytes have biomarkers such as CD56 and CD66b.

[0039] Based on the previous retrospective study of the research group, we included 30 patients with advanced melanoma of different subtypes (all patients had signed informed consent), with a median age of 50 years (23-69 years). All these patients received anti-PD-1 treatment. Other clinical characteristics are shown in Table 1. FFPE sections of melanoma were taken from patients for multiple immunofluorescence staining, targeting S100, CD68, CD3, CD20, PD-L1 and PD-1. The multicolor immunofluorescence kit was purchased from Absin and contains all the reagents required for the following operations:

[0040] 1. Dewaxing and hydration

[0041] a) Soak the slides in fresh xylene for 10 min, repeat 3 times.

[0042] b) Immerse the slides in gradient ethanol: 100% for 5 min; 95% for 5 min; 70% for 2 min.

[0043] c) Wash the slides with sterile water for 1 min and repeat 3 times.

[0044] d) Soak the slides in 10% neutral formalin or 4% paraformaldehyde for 10-30 minutes, wash the slides with sterile water for 1 minute, and repeat 3 times.

[0045] e) Add membrane permeabilization agent for 15 minutes, then rinse with PBST for 3 minutes, and repeat 3 times.

[0046] 2. Microwave repair of antigens

[0047] a) Place the dewaxed and hydrated slide in a repair cup and immerse it in 1× antigen repair solution.

[0048] b) Place the repair cup in a microwave oven and boil on high heat.

[0049] c) Maintain low heat for 15 minutes (pay attention to replenishing fluid to prevent excessive evaporation and causing dry slices).

[0050] d) Take out the container and allow to cool naturally to room temperature.

[0051] 3. Quenching and blocking

[0052] a) Remove the remaining wash solution on the slide.

[0053] b) Add 3% hydrogen peroxide to cover the sample area, incubate for 10 minutes, and then rinse with PBST for 3 minutes. Repeat this process 3 times.

[0054] c) Remove the remaining washing solution on the slide, add blocking solution to cover the sample area, keep moist and shake at room temperature for 10-30 minutes, and remove the blocking solution.

[0055] 4. Primary Antibody Incubation

[0056] a) Remove the blocking solution from the slides.

[0057] b) Use a pipette to add the diluted primary antibody solution dropwise to submerge the sample area.

[0058] c) Incubate with shaking for 1 h at room temperature.

[0059] d) Wash the slides with 1× TBST buffer for 3 min and repeat once.

[0060] 5. Secondary Antibody Incubation

[0061] a) Remove the remaining wash solution on the slide.

[0062] b) Directly add HRP secondary antibody working solution to immerse the sample area.

[0063] c) Incubate at room temperature for 10 minutes.

[0064] d) Wash the slides with 1× TBST buffer for 3 min. Repeat once.

[0065] 6. Fluorescence staining to amplify the signal

[0066] a) Remove the remaining wash solution on the slide.

[0067] b) Use a pipette to add 100ul of 1× dye working solution (diluted 1:100 with signal amplification solution) onto the slide to immerse the sample area.

[0068] c) Incubate with shaking for 10 min at room temperature.

[0069] d) Wash the slides with 1×TBST buffer and soak them at room temperature for 3 minutes. Repeat 3 times.

[0070] e) Microwave repair and naturally cool to room temperature.

[0071] f) Wash the slides once with sterile water and soak them in 1× TBST buffer for 2 min.

[0072] 7. New round of dyeing

[0073] a) After each round of staining, the staining condition can be confirmed by fluorescence microscopy. Be sure to cover the sample with TBST to prevent the slide from drying. b) Blocking: Remove the remaining wash solution on the slide, add a drop of blocking solution, cover the sample area, keep it moist and shake at room temperature for 10-30 minutes, and remove the blocking solution.

[0074] c) Repeat steps 4-6 several times until all markers are stained.

[0075] 8. Nuclear staining and sealing

[0076] Add 1×DAPI working solution to the sample, immerse the sample area, and incubate at room temperature for 5 minutes. Wash the slides 3 times with 1×TBST, 2 minutes each time. Then add anti-fluorescence quenching mounting agent and seal the slides with coverslips to avoid bubbles. For long-term storage, please seal the edges of the coverslips with transparent nail polish.

[0077] 9. Read the slides and observe and analyze the stained tissue slides under a fluorescence microscope.

[0078] 10. Image Analysis

[0079] The results of multiple immunofluorescence staining were scanned in full sections using a variety of commercial scanners. After that, the Deep-Learning algorithm based on the PhenoVision Pat-FL spatial phenotype analysis software was used to automatically divide each cell after training with tens of thousands of nuclei with different staining intensities. After that, the protein expression phenotype of each cell was identified by PhenoVision Pat-FL, the relevant parameters were formulated through positive controls, and an APP was established to perform data analysis on all sections of the same panel. On the basis of cell identification, the adjacent tissues were divided into two distance gradients of 0-25 and 25-50 microns, starting from tumor cells (S100 positive), and the number and percentage of CD3, CD20, CD68, PD-L1, PD-1, PD-1+CD68, PD-1+CD3, PD-1+CD20, PD-L1+CD68, PD-L1+CD3, PD-L1+CD20, and PD-L1 positive cells in tumor tissues were identified and quantitatively calculated.

[0080] 11. Follow-up records: The follow-up system records the immunotherapy response and survival of the above patients. According to the RECIST standard for evaluating the efficacy of solid tumors, the immune response is divided into: partial response (PR), that is, the sum of the maximum diameters of the target tumor lesions decreases by ≥30% and lasts for at least 4 weeks; progressive disease (PD), the sum of the maximum diameters of the target tumor lesions increases by at least ≥20%, or new lesions appear.

[0081] The staining results are as follows Figure 1 As shown: It can be seen that the expression levels of CD68, CD3, S100, CD20, PDL1 and PD1 can be simultaneously displayed on a paraffin section with good resolution. After tracking and following up the 30 patients enrolled in the above group, it was found that the median follow-up time of the entire cohort was 7.43 months from the start of anti-PD-1 treatment. In the entire cohort, 13 patients (43.33%) maintained PR during the follow-up period, including 4 acrofacial subtypes, 4 mucosal subtypes, and 5 skin subtypes. 17 patients (56.67%) experienced PD during the follow-up period, including 8 acrofacial subtypes, 5 mucosal subtypes, and 4 skin subtypes. The PFS of the PR group and PD group were 24.43 months and 2.37 months, respectively.

[0082] Table 1: Clinical and pathological characteristics of patients (N=30)

[0083]

[0084]

[0085] Abbreviation: NA, not applicable

[0086] We performed data analysis based on the signal recognition of the sample images obtained above and the subsequent bioinformatics analysis combined with the biological large cohort. In the analysis, we found that the two cell populations of PD-1 positive and CD3 / PD-1 co-positive were significantly different between the PD and PR group patients (p=0.0033 and p=0.0294). The proportion of PD-1 positive and CD3 / PD-1 co-positive cells in the PR group was significantly higher than that in the PD group (PD-1: PR group average = 0.3200%, PD group average = 0.05294%, p = 0.0033; CD3 / PD-1: PR group average = 0.1723%, PD group average = 0.02412%, p = 0.0294) ( Figure 1 C). In summary, the higher the proportion of PD1+ cells in the patient's tumor tissue, the better the effect of anti-PD-1 immunotherapy. At the same time, the more PD-1+ CD3+ T infiltration, the better the efficacy of anti-PD-1 immunotherapy. This is consistent with the obvious effect of immune checkpoint inhibitor therapy in patients with high PD1 expression. Therefore, these results highlight the value of studying specific tumor-infiltrating immune cells.

[0087] We also unexpectedly found that the spatial distance between tumor cells and other cells is also closely related to the efficacy of immunotherapy. The proportion of PD-1 positive cells and CD3 / PD-1 co-positive cells within the range of 0-25um of melanoma cells (S100+) was statistically significantly related to the efficacy of PD-1 treatment (p=0.0109 and p=0.0312). Compared with patients in the PD group, the proportion of PD-1 positive cells within 0-25μm of melanoma cells (S100+) in the PR group was significantly higher ( Figure 2 B) (PD-1: PR group mean = 0.3262%, PD group mean = 0.07176%, p = 0.0109; CD3 / PD-1: PR group mean = 0.1885%, PD group mean = 0.03471%; p = 0.0312). This shows that among the several cell subsets analyzed, the closer the infiltration of PD-1+ cells in the PR group and the distance to tumor cells (0-25um), the better the predicted efficacy of the patients.

[0088] In the range of 25-50 μm around melanoma cells (S100+), the proportions of PD-1-positive, PD-1 / CD3-co-positive and PD-1 / CD20-co-positive cells were significantly correlated with the efficacy of PD-1 therapy (P=0.0074, P=0.0297 and P=0.0157). In the PR group, the proportions of PD-1-positive, PD-1 / CD3-co-positive and PD-1 / CD20-co-positive cells within 25-50 μm from melanoma cells were significantly correlated with the efficacy of PD-1 therapy (P=0.0074, P=0.0297 and P=0.0157). The proportion of D20 co-positive cells was significantly higher than that of patients in the PD group (PD-1: PR group mean = 0.4454%, PD group mean = 0.07471%, p = 0.0074; PD-1 / CD3: PR group mean = 0.3200%, PD group mean = 0.03235%; p = 0.0297; PD-1 / CD20: PR group mean = 0.02769%, PD group mean = 0.003529%; p = 0.0157) Figure 2 C).

[0089] The above results indicate that within a certain distance, the higher the infiltration of PD-1-positive CD3+T cells and CD20+B cells around tumor cells, the better the effect of anti-PD-1 treatment.

[0090] Example 2 Construction of prediction model

[0091] According to the distribution characteristics of the six targets (S100, CD68, CD3, CD20, PD-L1 and PD-1) in the tumor area in multiple immunofluorescence, combined with the patient clinical cohort information (i.e., the clinical information of patients in the PD group and PR group who received PD-1 antibody treatment), a efficacy prediction model was constructed. Specifically, DAPI was used to identify the number of cells, and the overall positive rate of S100, CD68, CD3, CD20, PD-L1 and PD-1 single-positive marker cells, the distance of other cells from S100+ positive cells, the positive rate of PD1+CD3+ cells, the positive rate of PD1+CD20+ cells, the positive rate of immune cells (CD68+, CD3+, CD20+, PD-L1+ and PD-1+) within 25um around the center of S100+ positive cells; the positive rate of immune cells (CD68+, CD3+, CD20+, PD-L1+ and PD-1+) within 25um-50um around the center of S100+ positive cells were calculated.

[0092] Immunotherapy efficacy score = number of biomarker-positive cells / total number of DAPI-stained (S100, CD68, CD3, CD20, PD-L1, and PD-1)-positive cells;

[0093] or,

[0094] Immunotherapy efficacy score = number of biomarker-positive cells (cells within 0-25um around S100+ cells) / total number of DAPI-stained (S100, CD68, CD3, CD20, PD-L1 and PD-1)-positive cells (cells within 0-25um around S100+ cells)

[0095] or,

[0096] Immunotherapy efficacy score = number of biomarker-positive cells (cells within 25-50um around S100+ cells) / total number of DAPI-stained (S100, CD68, CD3, CD20, PD-L1 and PD-1)-positive cells (cells within 25-50um around S100+ cells)

[0097] The efficacy of immunotherapy in patients is evaluated based on the Cutoff value (predicted critical value) of the immunotherapy efficacy score corresponding to different biomarkers. Patients with values ​​higher than the cutoff value are judged as those who benefit from immunotherapy, and patients with values ​​lower than the cutoff value are judged as those who do not benefit from immunotherapy (ineffective).

[0098] Furthermore, we evaluated and calculated the ROC curve using a biostatistical method (Abinary logistic regression model) Figure 3 A).

[0099] We first generated the ROC curves of PD-1 single positive and CD3 / PD-1 co-positive markers based on the clinical response data of the enrolled patients (such as Figure 3 B, 3C), the response score AUC for PD-1 positivity was 0.8824 (95% CI: 0.722-1.0), and the response score AUC for PD-1 / CD3 co-positivity was 0.7624 (95% CI: 0.722-1.0). The cutoff values ​​of these two indicators were calculated as 0.12% and 0.05%, respectively.

[0100] Furthermore, we introduced the spatial distribution of cells into the model, and generated ROC curves for PD-1 single positive and CD3 / PD-1 co-positive markers in the range of 0-25um of tumor cells based on the clinical response data of patients (e.g. Figure 3D, 3E), the response score AUC for PD-1 positivity was 0.7964 (95% CI: 0.722-1.0), and the response score AUC for PD-1 / CD3 co-positivity was 0.6946 (95% CI: 0.4895-0.8996). The cutoff values ​​of these two markers within 0-25um calculated from this were 0.015% and 0.035%, respectively. Similarly, the ROC curve for tumor cells within the range of 25-50um was generated based on the patient's clinical response data (such as Figure 3 F, 3G, 3H), the reaction score AUCs for PD-1 positivity, PD-1 / CD3 co-positivity, and PD-1 / CD20 co-positivity were 0.8824 (95% CI: 0.7566 to 1.000), 0.8394 (95% CI: 0.6828 to 0.9959), and 0.6810 (95% CI: 0.4752 to 0.8868), respectively. The cut-off values ​​of the above indicators were calculated to be 0.095%, 0.055%, and 0.025%, respectively. It can be seen that when the spatial distribution of cells is added, the positive rate of PD-1 / CD3 co-positive cells significantly improves the predictive effect of melanoma immunotherapy.

[0101] Example 3: Validation of the efficacy of biomarkers in predicting the efficacy of melanoma immunotherapy

[0102] In order to detect the predictive effect of the marker combination of the present invention on the patient's immunotherapy response, 30 melanoma patients were re-enrolled (all patients have signed informed consent). These patients also received anti-PD-1 treatment, and their treatment prognosis was tracked. According to the patient's prognosis, they were divided into PR group and PD group. The melanoma FFPE sections of the patients were taken for multiple immunofluorescence staining targeting S100, CD68, CD3, CD20, PD-L1 and PD-1.

[0103] The results of staining analysis showed that the expression levels of PD1+ and PD1+CD3+ in PR patients were significantly higher than those in PD patients. It was further confirmed that a higher proportion of PD-1 positive and CD3 / PD-1 co-positive cell subsets was associated with a longer PFS (p=0.0016 and p=0.0021) ( Figure 3 B and C).

[0104] Using S100 as a tumor cell marker, the spatial relationship between PD1+ or PD1+CD3+ cells and tumor cells was recorded. If more than half of the PD1+ or PD1+CD3+ cells were less than 25um away from tumor cells, the patient was included in the PD1+_25μm group or PD1+CD3+_25μm group, otherwise the patient was included in the PD1-_25μm group or PD1+CD3-_25μm group. The response of patients in different groups to PD1 treatment was analyzed. The results showed that the immune response of patients in the PD1+_25μm and PD1+CD3+25μm groups was significantly better than that in the PD1-_25μm group and PD1+CD3-_25μm group. The PFS of patients with a higher proportion of positive cells was significantly higher than that of patients with a lower proportion (p=0.0003 and p=0.00719) ( Figure 3 D and E).

[0105] This scheme can also effectively predict the prognosis of patients with spatial distribution between 25-50μm. The relevant results are shown in Figure 3 F, G, and H show that the progression-free survival (PFS) of patients in the PD1+_25-50 group, PD1+CD3+_25-50 group, and PD1+CD20+_25-50 group was significantly better than that of patients in the PD1-_25-50 group, PD1+CD3-_25-50 group, and PD1+CD20-_25-50 group (p=0.0002, p=0.0004, and p=0.016).

[0106] Using PD-1 positivity as a marker for predicting efficacy, according to the immunotherapy efficacy scoring model described in Example 2, 19 patients were predicted to be PD, 11 patients were predicted to be PR, and 24 patients had the predicted efficacy that was completely consistent with the actual response of the patients, with a prediction accuracy of 80% (Table 2).

[0107] Table 2 Using PD-1 positivity as a marker to predict efficacy

[0108]

[0109] a. The cutoff value is .500

[0110] PD-1 / CD3 co-positivity was used as a marker to predict efficacy. Among them, 17 patients were predicted to be PD, 8 patients were predicted to be PR, and 23 patients had the predicted efficacy completely consistent with the actual response of the patients. The prediction accuracy rate reached 76.7% (Table 3).

[0111] Table 3 Using PD-1 / CD3 co-positivity as a marker to predict efficacy

[0112]

[0113] a. The cutoff value is .500

[0114] Using PD-1 positivity or PD-1+CD3+ double positive in the range of 0-25um tumor cells as markers for predicting efficacy, 21 and 22 cases were predicted to be PD, and 9 and 8 cases were predicted to be PR. The predicted efficacy and actual response were completely consistent in 22 and 21 patients, respectively, with prediction accuracy rates of 73.3% and 70%, respectively (Table 4-5).

[0115] Table 4 Using PD-1 positivity in the range of 0-25um on tumor cells as a marker for predicting efficacy

[0116]

[0117] a. The cutoff value is .500

[0118] Table 5 Using PD-1+CD3+ double positive in the range of 0-25um tumor cells as a marker to predict efficacy

[0119]

[0120] a. The cutoff value is .500

[0121] PD-1 positive, PD-1+CD3+ or PD-1+CD20+ positive markers within the range of 25-50um of tumor cells were used to predict 19, 20 and 22 patients as PD, and PR was predicted in 11, 10 and 8 cases, respectively. Among them, the actual efficacy and predicted efficacy were completely consistent in 22, 23 and 21 patients, respectively, with prediction accuracy rates of 73.3%, 76.7% and 70%, respectively (Table 6-8).

[0122] Table 6 Using PD-1 positivity in the range of 25-50 μm on tumor cells as a marker for predicting efficacy

[0123]

[0124] a. The cutoff value is .500

[0125] Table 7 Using PD-1+CD3+ double positive in the range of 25-50um of tumor cells as a marker to predict efficacy

[0126]

[0127] a. The cutoff value is .500

[0128] Table 8 Using PD-1+CD20+ double positive in the range of 25-50um of tumor cells as a marker to predict efficacy

[0129]

[0130] a. The cutoff value is .500

[0131] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. Use of a detection reagent in the preparation of a kit for predicting the efficacy of anti-PD-1 immunotherapy for melanoma, characterized in that: The detection reagent is an immunofluorescence reagent, and the biomarkers are CD68, CD3, S100, CD20, PDL1 and PD1.

2. The use according to claim 1, characterized in that The application includes using the immunofluorescence reagent to stain the patient's melanoma sample, and predicting the efficacy of anti-PD-1 immunotherapy for the patient by identifying and quantitatively calculating the number and percentage of CD3, CD20, CD68, PD-L1, PD-1, PD-L1, PD-1+CD68, PD-1+CD3, PD-1+CD20, PD-L1+CD68, PD-L1+CD3, and PD-L1+CD20 positive cells in tumor tissue.

3. The use according to claim 2, characterized in that The calculation includes calculating an immunotherapy prediction score, Immunotherapy efficacy score = number of biomarker-positive cells / total number of DAPI-stained (S100, CD68, CD3, CD20, PD-L1, and PD-1)-positive cells; or, Immunotherapy efficacy score = number of biomarker-positive cells (cells within 0-25 um around S100+ cells) / total number of DAPI-stained (S100, CD68, CD3, CD20, PD-L1 and PD-1)-positive cells (cells within 0-25 um around S100+ cells); or, Immunotherapy efficacy score = number of biomarker-positive cells (cells within 25-50 um around S100+ cells) / total number of DAPI-stained (S100, CD68, CD3, CD20, PD-L1 and PD-1)-positive cells (cells within 25-50 um around S100+ cells); The efficacy of immunotherapy in patients is evaluated based on the Cutoff value (predicted critical value) of the immunotherapy efficacy score corresponding to different biomarkers. Patients with values ​​higher than the cutoff value are judged as those who benefit from immunotherapy, and patients with values ​​lower than the cutoff value are judged as those who do not benefit from immunotherapy (ineffective).

4. The use according to claim 3, characterized in that Immunotherapy efficacy score = PD1+CD3+ cell number (cells within 25-50 um around S100+ cells) / DAPI-labeled cell number (cells within 25-50 um around S100+ cells), and the cutoff value is 0.

05.

5. A model for predicting the efficacy of anti-PD-1 immunotherapy for melanoma, characterized in that: The model is: Immunotherapy efficacy score = number of biomarker-positive cells / total number of DAPI-stained (S100, CD68, CD3, CD20, PD-L1, and PD-1)-positive cells; or, Immunotherapy efficacy score = number of biomarker-positive cells (cells within 0-25 um around S100+ cells) / total number of DAPI-stained (S100, CD68, CD3, CD20, PD-L1 and PD-1)-positive cells (cells within 0-25 um around S100+ cells); or, Immunotherapy efficacy score = number of biomarker-positive cells (cells within 25-50 um around S100+ cells) / total number of DAPI-stained (S100, CD68, CD3, CD20, PD-L1 and PD-1)-positive cells (cells within 25-50 um around S100+ cells); The model is used as follows: the immunotherapy efficacy of patients is evaluated according to the Cutoff value (prediction critical value) of the immunotherapy efficacy score corresponding to different biomarkers. Patients with values ​​higher than the cutoff value are judged as the group that benefits from immunotherapy, and patients with values ​​lower than the cutoff value are judged as the group that does not benefit from immunotherapy (ineffective).

6. The model according to claim 5, characterized in that Immunotherapy efficacy score = PD1+CD3+ cell number (cells within 25-50 um around S100+ cells) / DAPI-labeled cell number (cells within 25-50 um around S100+ cells), and the cutoff value is 0.

05.

7. A kit for predicting the efficacy of anti-PD-1 immunotherapy for melanoma, characterized in that: The kit comprises a detection reagent, wherein the detection reagent is an immunofluorescence reagent, and the biomarkers are CD68, CD3, S100, CD20, PDL1 and PD1.

8. The kit according to claim 7, characterized in that The kit is used as follows: the immunofluorescence reagent is used to stain the patient's melanoma sample, and the efficacy of anti-PD-1 immunotherapy for the patient is predicted by identifying and quantitatively calculating the number and percentage of CD3, CD20, CD68, PD-L1, PD-1, PD-L1, PD-1+CD68, PD-1+CD3, PD-1+CD20, PD-L1+CD68, PD-L1+CD3, and PD-L1+CD20 positive cells in the tumor tissue.

Images