An in vitro detection method and kit for screening SMARCA2 degrader or inhibitor responders based on the degree of SMARCA4 functional inactivation

Abstract

The application discloses an in vitro detection method and kit for screening SMARCA2 degradation agent or inhibitor responders based on the degree of SMARCA4 functional inactivation, and relates to the field of companion diagnosis. In view of the problem that the existing method for selecting patients according to whether there is an arbitrary SMARCA4 mutation has a low enrichment degree of the enrichment of the missense variation with a function retention, the application determines the function consequence category (at least divided into a truncated function loss category and a missense category) of the SMARCA4 gene variation of an ex vivo biological sample, and detects the expression level of the SMARCA4 expression product and compares the expression level with a reference threshold value, when the variation belongs to the truncated function loss category and / or the expression level is lower than the reference threshold value, the test object is determined as a response candidate of the SMARCA2 selective degradation agent or inhibitor. The application also provides the use of a corresponding detection kit, a reagent combination and a data processing method for non-diagnostic and non-treatment purposes in vitro, and the enrichment degree of the responders is increased from about 1.4% to about 33.3%. The abstract drawing is Fig. 1.

Description

Technical Field

[0001] This invention relates to the field of precision oncology and companion diagnostics, specifically to a method for detecting the degree of functional inactivation of SMARCA4 in an in vitro sample, a detection kit, the use of the reagent combination in the preparation of the detection kit, and an in vitro data processing method for non-diagnostic and non-therapeutic purposes, used to screen response candidates for selective degradation agents or inhibitors of SMARCA2 from test subjects. Background Technology

[0002] SMARCA2 (also known as BRM) and SMARCA4 (also known as BRG1) are two paralogous catalytic ATPase subunits in the SWI / SNF chromatin remodeling complex. They are highly homologous in structure and function and can compensate for each other in the complex. When SMARCA4 loses function in tumor cells, the cells develop a strong dependence on its paralogous subunit SMARCA2, a phenomenon known as synthetic lethality: in the context of SMARCA4 inactivation, further inhibition or degradation of SMARCA2 selectively kills tumor cells, while having less impact on cells with normal SMARCA4 function. Based on this synthetic lethality relationship, selective degraders or inhibitors targeting SMARCA2 have been developed for the treatment of tumors with SMARCA4 deficiency. The relevant tumor types mainly include solid tumors such as non-small cell lung cancer and esophageal cancer. According to literature reports, SMARCA4 deficiency is seen in about 5% to 10% of non-small cell lung cancer, and also in various solid tumors such as esophageal cancer, gastroesophageal junction cancer, and hypercalcemic small cell ovarian cancer.

[0003] At the drug development level, several selective SMARCA2 degraders or inhibitors are currently in clinical or preclinical research stages, demonstrating a trend of multiple research entities pursuing the same mechanism in parallel. For example, PRT3789 (Prelude Therapeutics) and FHD-909 (also known as LY4050784, Foghorn Therapeutics / Eli Lilly) have entered clinical trials. PRT3789 has observed objective responses meeting the Evaluation Criteria for Solid Tumor Treatment (RECIST) in patients with SMARCA4 deficiency in non-small cell lung cancer, esophageal cancer, and gastric cancer (data as of the 2024 European Society for Medical Oncology Congress). Given this situation, accurately screening for truly SMARCA2 synthesis-lethal dependent responses from cancer patients with SMARCA4 abnormalities, who can benefit from these drugs, has become a key issue in the development and clinical application of such drugs. Accurate patient selection testing methods not only determine the quality of clinical trial enrollment but also the precision of post-marketing administration.

[0004] The existing patient-selectable technology options mainly fall into the following two categories.

[0005] The first type of approach is a patient selection protocol based on nucleic acid sequencing and the "any SMARCA4 mutation" method. This protocol uses a next-generation sequencing (NGS) panel to detect the presence of SMARCA4 gene mutations in tumor samples. If a SMARCA4 mutation is detected, the patient is included as a candidate for treatment with a SMARCA2 selective degrader or inhibitor, without distinguishing the functional consequences of the mutation. This is a commercially available protocol that uses the "presence or absence of mutation" as the sole criterion; its technical implementation relies solely on mutation detection itself and does not further determine whether the mutation actually impairs the function of the SMARCA4 protein.

[0006] The second type of approach is a patient selection approach based on immunohistochemistry (IHC) for protein expression loss. This approach uses immunohistochemistry to detect whether SMARCA4 (BRG1) protein expression is absent in the sample, using protein expression loss as the sole selection indicator. This approach acknowledges that SMARCA4 protein expression loss is a readout of its functional inactivation, but it treats the presence or absence of protein expression as a single binary indicator, without combining it with the classification of the functional consequences of the variant.

[0007] Furthermore, as the closest prior art to this invention, US11865114B2 (applicant: Aurigene Oncology Limited, priority date: September 12, 2019, authorized: January 9, 2024) discloses a patient selection and treatment method for SMARCA2 / 4 degrading agents. The technical approach of this method is as follows: isolate samples from the subject, and detect the presence of at least one of three tumor-specific alterations in the sample: androgen receptor (AR) mutations, amplifications, or overexpressions; PTEN loss of function or harmful mutations; and TMPRSS2-ERG translocations. If the presence of any of these alterations is detected, the subject is determined to be a responder to the SMARCA2 / 4 degrading agent, and the agent is administered accordingly. Its indication is prostate cancer. The prior art differs fundamentally from the present invention in terms of the biomarker, the criteria for determination, and the type of the object: First, the prior art focuses on downstream or bypass biomarkers in the context of prostate cancer, such as AR, PTEN, and TMPRSS2-ERG, rather than the functional state of SMARCA4 itself, while the present invention directly uses the degree of functional inactivation of SMARCA4 upstream of the target of SMARCA2 / 4 degraders as the criterion for determination; Second, the prior art only detects the "presence" of the corresponding changes, without distinguishing the functional consequences of the mutations or introducing expression levels as a continuous proxy for the degree of functional inactivation; Third, the prior art is a treatment method that includes a drug administration step and is limited to prostate cancer, while the present invention is an in vitro detection method that does not include a drug administration step and is applicable to solid tumors with SMARCA4 function loss, such as non-small cell lung cancer and esophageal cancer.

[0008] The existing technical solutions described above have the following technical disadvantages, and these disadvantages are precisely the technical problems that this invention aims to solve.

[0009] First, the presence or absence of a SMARCA4 mutation is not equivalent to SMARCA4 protein inactivation or SMARCA2 synthesis lethality. The first type of commercially available protocols conflate these three levels, classifying a sample as positive simply because a SMARCA4 mutation is detected. However, the presence of a mutation does not necessarily mean that SMARCA4 protein function is impaired, much less that the sample is in a synthesis lethal dependence on SMARCA2.

[0010] Secondly, the functional consequences of SMARCA4 gene variants vary greatly. Truncated variants such as nonsense mutations, frameshift mutations, and splice site mutations typically lead to loss of protein function, while missense mutations and full-frame mutations often do not impair protein function, and their dependence on SMARCA2 is comparable to that of the wild type. The first type of commercially available treatments fail to distinguish between these two distinct categories of variants, treating functionally preserving missense / full-frame variants and functionally inactivating truncated variants interchangeably. This results in the misclassification of many patients who are not dependent on SMARCA2 and do not respond to these drugs as treatment candidates, leading to low responder enrichment and a decline in the quality of clinical trial enrollment.

[0011] Third, the SMARCA4 gene mutation status and the loss of BRG1 protein expression are not entirely consistent; each has its blind spots. On the one hand, literature reports that about half of the samples carrying the SMARCA4 mutation do not show the loss of BRG1 protein expression, meaning the presence of the mutation does not equate to functional inactivation. On the other hand, there are also samples where immunohistochemistry confirms the loss of BRG1 protein expression, but NGS mutation detection does not detect the corresponding mutation (e.g., loss of heterozygosity, large fragment deletion, and other non-point mutation-induced functional inactivation). Such samples may be missed if the mutation is detected solely. Therefore, the determination of functional inactivation requires a joint assessment of the functional consequences of the mutation and the expression level; using either method alone may misidentify true responders dependent on SMARCA2.

[0012] Fourth, the second type of scheme that simply detects the loss of SMARCA4 protein expression uses protein expression as an isolated binary indicator without combining it with the functional classification of variants. It is difficult to quantify the continuously changing degree of functional inactivation into an operable and stratified response prediction indicator, and it is also impossible to make a comprehensive judgment on samples that have both truncated variants and reduced expression.

[0013] Existing technologies either rely solely on the presence or absence of mutations without considering functional consequences, or only on protein expression loss without considering mutation classification. Neither of these methods can accurately quantify the degree of functional inactivation of SMARCA4 and thus enrich truly SMARCA2-dependent responders. There is an urgent need in this field for an in vitro detection method and corresponding kit that uses the degree of functional inactivation of SMARCA4 (rather than the presence or absence of mutations) as a basis, combines the category of functional consequences of mutations with continuous proxying of expression levels, and thereby accurately enriches responders to selective degraders or inhibitors of SMARCA2. Summary of the Invention

[0014] To address the shortcomings of the existing technologies, the present invention aims to provide an in vitro detection method, corresponding detection kit, uses, and data processing method for screening SMARCA2 selective degradation agents or inhibitor response candidates based on the degree of functional inactivation of SMARCA4. This method aims to eliminate false positives caused by functionally preserved variants and compensate for missed detections in simple variant detection. As a result, it accurately enriches truly SMARCA2 synthesis-lethal dependent responders from subjects carrying SMARCA4 abnormalities, thereby improving the accuracy of clinical trial enrollment and medication.

[0015] To achieve the above objectives, the core technical concept of this invention is: instead of using "whether any SMARCA4 mutation is detected" as the screening criterion, the screening criterion is the degree of functional inactivation of SMARCA4, which is characterized by two dimensions: first, the functional consequence category of the SMARCA4 gene mutation; and second, the expression level of the SMARCA4 expression product. The combination of these two dimensions constitutes the technical feature that distinguishes this invention from the prior art.

[0016] Therefore, the present invention provides the following four technical solutions, which are based on the same general inventive concept and serve as backups for each other.

[0017] In a first aspect, the present invention provides a method for detecting the degree of functional inactivation of SMARCA4 in an in vitro sample, used to screen response candidates for selective degraders or inhibitors of SMARCA2 from test subjects, comprising: determining the functional consequence category of SMARCA4 gene variants in an in vitro biological sample from the test subject, the functional consequence category including at least a truncated loss-of-function category and a missense category, wherein the truncated loss-of-function category includes nonsense mutations, frameshift mutations, and splice site mutations, and the missense category includes missense mutations and whole-frame mutations; detecting the expression level of SMARCA4 expression products in the in vitro biological sample and comparing the expression level with a reference threshold; determining the degree of functional inactivation of SMARCA4 based on the functional consequence category and the expression level, wherein when the SMARCA4 gene variant belongs to the truncated loss-of-function category, and / or the expression level is lower than the reference threshold, the test subject is identified as a response candidate for a selective degrader or inhibitor of SMARCA2; wherein the selective degrader or inhibitor of SMARCA2 is a response candidate for selective degraders or inhibitors of SMARCA2. Any selective degradation agent or inhibitor, not limited to a specific compound.

[0018] In this invention, the truncated loss-of-function category can be further defined as a variant annotated as HIGH impact or LikelyLoF based on variant effect prediction; the missense category is determined to be functionally preservable, and when it exists alone without being accompanied by an expression level below a reference threshold, it is not included in the criteria for identifying the subject as a response candidate. This additional technical feature ensures that functionally preserving variants are not mistakenly counted as response candidates, thereby eliminating false positive sources in existing commercial solutions.

[0019] In this invention, the expression level of the SMARCA4 expression product is the expression level of SMARCA4 mRNA or the expression level of SMARCA4 protein; the reference threshold is a threshold set relative to the expression level distribution of a reference population of SMARCA4 that is not functionally inactivated, and the expression level being lower than the reference threshold indicates an enhanced degree of functional inactivation of SMARCA4.

[0020] In this invention, determining the functional consequence category of the SMARCA4 gene mutation can be achieved by performing nucleic acid sequencing on the in vitro biological sample and annotating the functional effect of the detected mutation; detecting the expression level of the SMARCA4 expression product can be achieved by RNA quantification or immunohistochemistry. The test subjects can be individuals with non-small cell lung cancer or esophageal cancer.

[0021] In this invention, the method may further include: determining a SMARCA2 function dependence score based on the functional consequence category and the expression level, and using the SMARCA2 function dependence score, the functional consequence category, and the expression level together for determining the response candidates, wherein the SMARCA2 function dependence score is a continuous value characterizing the degree of dependence of the sample on SMARCA2. The SMARCA2 selective degrader may be a selective SMARCA2 / BRM protein degradation targeting chimeric (PROTAC) type degrader.

[0022] Secondly, the present invention provides a detection kit for screening candidates for selective degradation agents or inhibitors of SMARCA2, comprising: reagents for determining the functional consequence category of SMARCA4 gene mutations, wherein the functional consequence category includes at least a truncated loss-of-function category and a missense category; and reagents for measuring the expression level of SMARCA4 expression products; the detection kit is used to implement the method described in the first aspect. The reagents for determining the functional consequence category of SMARCA4 gene mutations may be nucleic acid sequencing reagents; the reagents for measuring the expression level of SMARCA4 expression products may be RNA quantification reagents or SMARCA4 protein immunohistochemical reagents; the detection kit may further include a control or internal control for determining the reference threshold.

[0023] Thirdly, the present invention provides a combination of reagents for determining the functional consequence category of SMARCA4 gene variants and reagents for measuring the expression level of SMARCA4 expression products, used in the preparation of a detection kit for screening response candidates of SMARCA2 selective degraders or inhibitors, wherein the functional consequence category includes at least a truncated loss-of-function category and a missense category, the truncated loss-of-function category including nonsense mutations, frameshift mutations, and splice site variations; the detection kit is configured to indicate that the corresponding test subject is a response candidate of SMARCA2 selective degraders or inhibitors when the SMARCA4 gene variant belongs to the truncated loss-of-function category and / or the expression level of the SMARCA4 expression product is below a reference threshold.

[0024] Fourthly, the present invention provides an in vitro, non-diagnostic, and non-therapeutic data processing method, comprising: acquiring SMARCA4 gene variant annotation data and SMARCA4 expression product expression level data of a sample; determining, based on the variant annotation data, whether the SMARCA4 variant of the sample belongs to the truncated loss-of-function category, wherein the truncated loss-of-function category includes nonsense mutations, frameshift mutations, and splice site mutations, or variants annotated as HIGH impact or LikelyLoF based on variant effect prediction; comparing the expression level data with a reference threshold; and outputting a molecular typing identifier based on the above results, wherein the molecular typing identifier characterizes the degree of functional inactivation of SMARCA4 in the sample, including whether it belongs to the truncated loss-of-function category and whether the expression level is lower than the reference threshold; the output of the data processing method is limited to the molecular typing identifier, and does not output disease diagnosis conclusions, nor does it include drug administration or treatment steps.

[0025] The beneficial effects of this invention are as follows.

[0026] By categorizing SMARCA4 gene variants into two functional consequence categories (truncated loss-of-function category and missense category), and using the expression level of SMARCA4 expression products as a continuous proxy for the degree of functional inactivation, this invention combines the two to determine the degree of SMARCA4 functional inactivation. This allows for the enrichment of samples with true functional inactivation and synthetic lethal dependence on SMARCA2 from the population carrying SMARCA4 abnormalities. Based on proprietary in-silico proof-of-concept (PoC) data, compared to existing commercially available patient selection schemes using "any SMARCA4 mutation," this invention increases the enrichment of responders from approximately 1.4% of the entire population in the heterogeneous "any mutation" group to approximately 33.3% in the truncated loss-of-function category (see Specific Implementation Example 2 for details).

[0027] By classifying missense categories as potentially functionally preserved and excluding them from the response candidate selection when they exist alone, this invention avoids the misclassification of missense mutations that are not significantly different from wild-type in SMARCA2 dependence, thus eliminating a major source of false positives in existing commercial solutions.

[0028] By introducing SMARCA4 expression level as a continuous functional inactivation proxy indicator, this invention can further stratify beyond the variant functional category, and can also identify functional inactivation samples that show significantly reduced expression, including non-point mutations, thus making up for the missed detection of such samples by simple variant detection.

[0029] Furthermore, all four aspects of the technical solutions of this invention are described as in vitro detection methods, detection kits (products), the use of reagent combinations in the preparation of kits, and data processing methods for in vitro non-diagnostic and non-therapeutic purposes. The output of the data processing methods is limited to the intermediate result of molecular subtyping identification, thereby circumventing the provisions of Article 25 of the Patent Law regarding the non-patentability of "methods for the diagnosis and treatment of diseases". Attached Figure Description

[0030] Figure 1 is a schematic diagram of the overall process of the in vitro detection method of the present invention.

[0031] Figure 2 is a logical diagram of the classification of variant functional consequences and comparison of expression thresholds to determine response candidates in this invention.

[0032] Figure 3 is a schematic diagram of the SMARCA2 dependency hierarchy of the three types of samples in this invention: wild type, missense category, and truncated loss of function category.

[0033] Figure 4 is a schematic diagram of the module structure of the detection kit and data processing system of the present invention.

[0034] The markings in each of the attached figures are explained as follows:

[0035] S1 represents the steps of obtaining in vitro biological samples and extracting nucleic acids and expression targets; S2 represents the steps of detecting SMARCA4 gene variants and annotating their functional effects; S3 represents the steps of determining the functional consequence category of SMARCA4 gene variants; S4 represents the steps of detecting the expression level of SMARCA4 expression products; S5 represents the steps of comparing the expression level with a reference threshold; S6 represents the steps of determining the degree of functional inactivation of SMARCA4 based on the functional consequence category and expression level and identifying response candidates; S7 represents the steps of optionally determining the SMARCA2 function dependence score and incorporating it into the determination; S8 represents the steps of outputting the response candidate determination results or molecular subtyping identifiers.

[0036] 201 indicates the truncated loss of function category decision branch; 202 indicates the missense category decision branch; 203 indicates the expression level is below the reference threshold decision branch; 204 indicates the expression level is not below the reference threshold decision branch; 205 indicates the output result determined as a response candidate; 206 indicates the output result determined as a non-response candidate.

[0037] 301 represents the wild-type sample group; 302 represents the missense category sample group; 303 represents the truncated loss-of-function category sample group; 304 represents the SMARCA2 strong dependency decision boundary.

[0038] M1 represents the sample receiving and nucleic acid extraction module; M2 represents the variant detection and functional effect annotation module; M3 represents the variant functional consequence classification module; M4 represents the expression level determination module; M5 represents the expression level and reference threshold comparison module; M6 represents the comprehensive determination module of functional inactivation degree; M7 represents the SMARCA2 function dependence scoring module; M8 represents the result output module. Detailed Implementation

[0039] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the following embodiments are only used to illustrate the present invention and not to limit the scope of protection of the present invention. Equivalent substitutions and adaptive adjustments made by those skilled in the art based on the following embodiments without departing from the concept of the present invention all fall within the scope of protection of the present invention.

[0040] The terminology used in this specification is as follows and is used consistently throughout. The term "truncated loss-of-function category" refers to the variant category that causes truncation or loss of function in the SMARCA4 protein, including nonsense mutations (stop-gained), frameshift mutations, and splice site mutations, or variants annotated as HIGHimpact or LikelyLoF based on variant effect prediction; these are also referred to as the first variant category herein. The term "missense category" refers to missense mutations and inframe mutations, whose function may be preserved; these are also referred to as the second variant category herein. The term "SMARCA4 expression product level" refers to the amount of SMARCA4 expression product (i.e., SMARCA4 mRNA or SMARCA4 protein), serving as a continuous proxy indicator of the degree of SMARCA4 inactivation. The term "reference threshold" refers to a preset threshold used to determine whether the SMARCA4 expression level is at a low expression level. The term "SMARCA4 functional inactivation degree" refers to an indicator reflecting the extent of SMARCA4 function impairment, characterized by both the category of functional consequences of mutations and the expression level. The term "response candidate" refers to subjects deemed suitable for treatment with SMARCA2 selective degraders or inhibitors by this method, representing an intermediate subtyping result. The term "molecular subtyping identifier" refers to an intermediate result identifier output by the data processing method, characterizing the degree of SMARCA4 functional inactivation in the sample, distinct from disease diagnostic conclusions. The term "SMARCA2 selective degrader or inhibitor" refers to any degrader or inhibitor selective for SMARCA2 (BRM), constituting the genus to which this invention applies, and is not bound to a single compound.

[0041] Unless otherwise specified, the sample processing, nucleic acid extraction, sequencing, variant annotation, expression quantification, immunohistochemistry, and other operations involved in the various embodiments of the present invention can all be achieved using conventional reagents, instruments, and procedures in the field. Those skilled in the art can reproduce these operations without creative effort based on the contents of this specification.

[0042] Example 1: Complete in vitro detection method flow

[0043] This embodiment provides a complete method for detecting the degree of functional inactivation of SMARCA4 in in vitro samples and screening candidate SMARCA2 selective degraders or inhibitors. As shown in Figure 1, the method includes steps S1 to S8; the determination logic for classifying the functional consequences of mutations and comparing expression thresholds is shown in Figure 2.

[0044] Step S1: Obtain in vitro biological samples and prepare test subjects. Obtain in vitro biological samples from the test subjects. These samples may be tumor tissue samples (including fresh frozen tissue, formalin-fixed paraffin-embedded tissue, i.e., FFPE tissue), fine-needle aspiration samples, or body fluid samples containing tumor-derived nucleic acids. Genomic DNA is extracted from these samples for variant detection; simultaneously, depending on the selected expression detection method, total RNA is extracted for mRNA quantification, or tissue sections are retained for protein immunohistochemical detection. For FFPE samples, routine pretreatment such as dewaxing and proteinase K digestion can be performed before nucleic acid extraction. The test subjects obtained in this step include the genomic DNA of the variant to be tested, and the RNA or tissue sections to be expressed.

[0045] Step S2 involves detecting and annotating functional effects of SMARCA4 gene variants. The genomic DNA obtained in Step S1 is subjected to nucleic acid sequencing, which can be next-generation sequencing targeting the SMARCA4 gene region, tumor multigene panel sequencing covering SMARCA4, or whole-exome sequencing. The detected SMARCA4 region variants are then detected and annotated. The annotation includes, but is not limited to, the genomic location of the variant, the reference and variant bases, the type of protein sequence change caused, and the functional impact level given by a variant effect prediction tool. In this embodiment, the variant effect annotation can adopt the variant effect prediction annotation process commonly used in the art. Its output includes at least the sequence ontology consequence type of the variant (such as stop_gained, frameshift_variant, splice_donor_variant, splice_acceptor_variant, missense_variant, inframe_insertion, inframe_deletion, etc.) and the corresponding functional impact level (such as HIGH, MODERATE, LOW), and may be attached with a functional loss probability indicator such as LikelyLoF.

[0046] Step S3: Determine the functional consequence category of SMARCA4 gene variants. Based on the annotation results obtained in Step S2, the detected SMARCA4 variants are classified into functional consequence categories, which include at least truncated loss-of-function and missense categories. Specifically, variants annotated as nonsense mutations (stop_gained), frameshift mutations, or splice site mutations are classified into the truncated loss-of-function category; variants annotated as missense mutations or inframe mutations (insertions or deletions) are classified into the missense category. In a preferred embodiment, the functional impact level given by the variant effect prediction tool can be directly used as the classification basis, that is, variants annotated as HIGHimpact or identified as LikelyLoF are classified into the truncated loss-of-function category. When a sample carries multiple SMARCA4 variants, as long as at least one of the variants belongs to the truncated loss-of-function category, the sample is considered to meet the truncated loss-of-function category criteria in the variant dimension.

[0047] As shown in Figure 2, this step corresponds to the following branches in the decision logic: 201 for truncated loss-of-function category and 202 for missense category. Samples belonging to the truncated loss-of-function category enter branch 201; samples carrying only missense variants enter branch 202. For samples carrying only missense variants without expression levels below the reference threshold, this method classifies them as potentially functionally preserved and does not include these missense variants in the criteria for determining the subject as a response candidate. This treatment is one of the key differences between this invention and existing commercial solutions: existing solutions determine a positive result as soon as any SMARCA4 mutation is detected, while this method excludes functionally preserved missense variants from the response candidate determination criteria, thereby avoiding false positives.

[0048] Step S4: Detect the expression level of the SMARCA4 expression product. The expression level of the SMARCA4 expression product is determined on the RNA or tissue sections obtained in Step S1. In embodiments using mRNA as the expression product, RNA quantification methods such as real-time quantitative PCR, digital PCR, hybridization-based expression quantification, or RNA sequencing can be used to determine the relative or absolute expression level of SMARCA4 mRNA, and normalized to the expression distribution of internal reference genes or the overall sample. In embodiments using protein as the expression product, immunohistochemistry (IHC) can be used to stain the SMARCA4 (BRG1) protein in the tissue sections, and the protein expression level is semi-quantitatively scored based on staining intensity and positive ratio.

[0049] Step S5: Compare the expression level of the SMARCA4 expression product with a reference threshold. The reference threshold is a threshold set relative to the expression level distribution of a non-inactivated SMARCA4 reference population. In one embodiment, a low quantile (e.g., the lower quartile or lower) of the expression level distribution of a non-inactivated SMARCA4 reference population (e.g., a wild-type sample population) can be used as the reference threshold; in another embodiment, a significant decrease relative to the median of the reference population (e.g., a decrease to below a certain proportion of the median) can be set as below the reference threshold. In the implementation using mRNA as the expression product, the SMARCA4 mRNA expression level of the test sample, after being normalized by an internal control, is compared with the expression distribution established by the reference population. If the expression level is lower than the value corresponding to the set quantile, it is judged to be below the reference threshold. In the implementation using protein as the expression product and employing immunohistochemistry, the semi-quantitative score of SMARCA4 protein staining (such as H-score or a combination score of staining intensity and positive ratio) can be judged to be below the reference threshold if it is lower than the critical score established by the reference population. In particular, the extreme case of complete absence of protein staining corresponds to an expression level significantly lower than the reference threshold. The reference threshold can be established during the kit calibration stage or the detection implementation stage, using the control or internal control described in Example 3. When the SMARCA4 expression level of the test sample is lower than the reference threshold, it indicates an enhanced degree of functional inactivation of SMARCA4 in the sample, and enters the expression level below the reference threshold determination branch 203; otherwise, it enters the expression level not lower than the reference threshold determination branch 204.

[0050] Step S6: Determine the degree of SMARCA4 functional inactivation based on the functional consequence category and expression level, and identify response candidates. Combining the results of the variation functional consequence category obtained in Step S3 and the expression level obtained in Step S5, the following logic applies: When the SMARCA4 gene variation belongs to the truncated loss-of-function category (branch 201), and / or the expression level is below the reference threshold (branch 203), the subject is identified as a response candidate for a selective degrader or inhibitor of SMARCA2, corresponding to output result 205; when the sample neither carries a truncated loss-of-function variation nor has an expression level below the reference threshold (i.e., only falls into the combination of branches 202 and 204), it is identified as a non-response candidate, corresponding to output result 206. That is, either a truncated loss-of-function category or low expression is sufficient to include the sample in the response candidate list; both conditions being met enhance the confidence in determining the degree of SMARCA4 functional inactivation, while those carrying only missense variation with high expression are excluded from the response candidate list.

[0051] Step S7: Optionally, the SMARCA2 functional dependency score is determined and included in the decision. As a preferred implementation, the SMARCA2 functional dependency score can be further calculated based on the variant functional consequence category and expression level. The SMARCA2 functional dependency score is a continuous value characterizing the sample's dependence on SMARCA2. It can be obtained by weighting the variant functional consequence category (e.g., assigning higher weights to truncated loss of function and lower weights to missense categories) and the normalized SMARCA4 expression level (assigning higher values ​​to lower expression levels) according to preset weights, or by combining other features reflecting the SMARCA4 functional state. Using the obtained SMARCA2 functional dependency score, variant functional consequence category, and expression level together for the decision of response candidates provides a continuous dependency hierarchy beyond binary decision-making, facilitating further ranking and enrichment within the response candidate group.

[0052] In one specific implementation, the SMARCA2 functional dependence score D can be calculated using the following formula: D = w1 · C + w2 · (1 − E_norm), where C is the assigned value for the variant functional consequence category, with C=1 for truncated loss of function, C=0 for missense, and C=0 for wild-type; E_norm is the normalized value of the SMARCA4 expression level of the test sample relative to the expression distribution of the SMARCA4 non-inactivated reference population, with a normalized value ranging from 0 to 1, where lower expression results in a smaller E_norm and a larger (1 − E_norm); w1 and w2 are preset weights, satisfying w1 + w2 = 1, for example, w1=0.6 and w2=0.4, so that both the variant functional category and expression level contribute to the score, with the variant category slightly dominating. The resulting score D ranges from 0 to 1, with a larger D indicating a stronger dependence of the sample on SMARCA2. The response candidates can be sorted from highest to lowest by D, or a scoring threshold (e.g., D not less than 0.5) can be set as an auxiliary condition for further enriching the response candidates. The above weight values ​​and scoring methods are only examples. Those skilled in the art can adaptively calibrate the weights and thresholds according to the actual queue. Any continuous dependency scoring that uses the category of variant functional consequences and the expression level as continuous proxies falls within the scope of this invention.

[0053] Step S8: Output the response candidate determination result or molecular subtyping identifier. Output the determination result obtained in step S6 (and optional step S7). In the embodiment as an in vitro detection method, the output is the determination result of whether the test subject belongs to the SMARCA2 selective degrader or inhibitor response candidate; in the embodiment as an in vitro data processing method for non-diagnostic and non-therapeutic purposes (see Example 3), the output is limited to a molecular subtyping identifier characterizing the degree of functional inactivation of SMARCA4 in the sample, that is, a combination of identifiers of "whether it belongs to the truncated loss of function category" and "whether the expression level is lower than the reference threshold", without outputting a disease diagnosis conclusion, nor including administration or treatment steps.

[0054] The SMARCA2 selective degrader or inhibitor described in this embodiment is any degrader or inhibitor selective for SMARCA2, and is not limited to a specific compound; it can be a selective SMARCA2 / BRM protein degradation targeting chimeric (PROTAC) type degrader, or a selective SMARCA2 / BRM small molecule inhibitor. The determination logic of this method does not depend on the specific chemical structure of the drug used, and therefore is applicable to any member within the drug genus.

[0055] The subjects described in this embodiment are preferably those with solid tumors that are prone to SMARCA4 dysfunction, such as those with non-small cell lung cancer or esophageal cancer.

[0056] Through the steps S1 to S8 above, this method replaces the existing "presence or absence of arbitrary mutations" as the screening criterion with the degree of functional inactivation of SMARCA4 (characterized by the category of functional consequences of mutations and the expression level). This not only eliminates false positives caused by functionally preserved missense variants, but also captures functional inactivation caused by non-point mutations through continuous expression proxy, thereby accurately enriching truly SMARCA2-dependent response candidates.

[0057] Example 2: Validation based on proprietary in-silico proof-of-concept data

[0058] This embodiment provides proprietary in-silico proof-of-concept (PoC) data supporting the feasibility and beneficial effects of the method in Example 1. This data is based on a retrospective computational analysis of a publicly available cohort, sufficient to demonstrate the feasibility of the judgment logic of "screening SMARCA2-dependent responders based on SMARCA4 variant functional consequence category + expression level," and is used to determine the application date. Wet experimental performance verification (such as detection limits and precision for detection platform migration, paired protein verification, etc.) serves as pre-authorization reinforcement and does not affect the feasibility of this method. The stratified relationship between the sample groups of each variant category and the SMARCA2 strong dependence judgment threshold is shown in Figure 3.

[0059] Cohort and Data Sources. This PoC utilizes DepMap data resources, integrating its 26Q1 CRISPR gene effect data with the 24Q4 SMARCA4 somatic mutation annotation data and accompanying gene expression data. The total number of cell lines included in the analysis after integration was n=1192. Based on the functional consequence category of the SMARCA4 variant carried by each cell line, the samples were divided into three groups: SMARCA4 wild-type (WT, i.e., not carrying the SMARCA4 somatic mutation) group (1052 cases); truncated loss-of-function category (i.e., Class I LoF) group (60 cases); and missense category (i.e., Class II missense) group (80 cases).

[0060] Criteria for determining SMARCA2 dependence. The degree of cell line dependence on SMARCA2 is measured by the CRISPR gene effect. A lower gene effect value indicates greater cell damage after SMARCA2 knockout, meaning a stronger dependence on SMARCA2. A SMARCA2 CRISPR gene effect value less than -0.5 is defined as strong dependence on SMARCA2, corresponding to the strong dependence threshold 304 in Figure 3.

[0061] The proportion of strong dependence increased with the degree of functional inactivation. In the three groups of samples mentioned above, the proportion of samples determined to be strongly dependent on SMARCA2 increased significantly with the degree of variant functional inactivation: 1.4% in the wild-type group (301); 7.5% in the missense group (302); and 33.3% in the truncated loss-of-function group (303). The proportion of strong dependence in the truncated loss-of-function group was approximately 24 times richer than that in the wild-type group. This result directly supports the core judgment logic of this invention: samples carrying truncated loss-of-function variants showed significantly higher dependence on SMARCA2 than wild-type and missense group samples, and therefore represent a highly enriched candidate group for selective degradation agents or inhibitors of SMARCA2.

[0062] Statistical tests support the non-obviousness of functional category differentiation. The Mann-Whitney test was used to examine the differences in SMARCA2 dependence among the groups. The results are as follows: there was a significant difference between the truncated loss-of-function group and the wild-type group (p=7.8e-9); a significant difference between the truncated loss-of-function group and the missense group (p=2.1e-4); while there was no significant difference between the missense group and the wild-type group (p=0.071). The result that there was no significant difference between the missense group and the wild-type is particularly crucial: it indicates that missense variants are comparable to wild-type in SMARCA2 dependence and do not constitute a valid basis for response enrichment, thus demonstrating from the data that missense variants are a source of false positives in existing commercial "arbitrary SMARCA4 mutation" schemes; at the same time, it shows that differentiating variants by functional consequence category (rather than simply statistically classifying them as "presence or absence of mutation") has a non-obvious and substantial effect compared to existing technologies.

[0063] The effectiveness of expression level as a continuous inactivation surrogate was investigated. Within a subset of samples carrying the SMARCA4 variant (n=137; 3 out of 140 variant-carrying cell lines were excluded due to lack of expression data), the relationship between SMARCA4 expression level and SMARCA2 dependence was examined. The results showed a significant correlation: Spearman correlation coefficient ρ=0.446, p=4.8e-8 (the correlation direction is expressed as "lower expression level, stronger SMARCA2 dependence," meaning that SMARCA4 expression level is negatively correlated with SMARCA2 dependence, corresponding to the positive correlation of CRISPR gene effects increasing with increasing expression; these are different expressions of the same relationship). The results support using SMARCA4 expression level as a continuous proxy indicator of the degree of functional inactivation—the lower the expression level, the stronger the degree of functional inactivation and the stronger the dependence on SMARCA2; thus supporting the present invention to use expression level below a reference threshold as another independent criterion for determining response candidates in addition to the variant functional category, and to identify functional inactivation samples that are not caused by point mutations and show reduced expression.

[0064] Quantification of Beneficial Effects. Based on the above data, compared with existing commercial patient selection methods for "any SMARCA4 mutation," this invention increases the responder enrichment from approximately 1.4% of the entire population (a strong dependence percentage in the wild-type group, approximately representing the background level before functional screening) to approximately 33.3% of the truncated loss-of-function group. If the existing commercial method includes all individuals carrying any SMARCA4 mutation (including missense categories), the strong dependence percentage of missense categories is only 7.5%, and there is no significant difference from the wild-type, which would significantly dilute the responder enrichment and incorrectly include a large number of non-responders. However, this invention, by excluding missense categories from the sole criterion and supplementing it with expression continuity proxy, can maintain the screening enrichment at approximately 33.3% of the truncated loss-of-function group, significantly reducing the incorrect inclusion rate of non-responders and improving the quality of clinical trial enrollment and the accuracy of medication.

[0065] Methods for setting the reference threshold and parameter sensitivity. Based on the above PoC data, the reference threshold can be set as follows, and exhibits corresponding parameter sensitivity. In the embodiment where the reference threshold is set based on the SMARCA4 expression level distribution of a wild-type sample population (i.e., a reference population with non-inactivated SMARCA4 function), a low quantile of the expression distribution of this reference population can be taken as the reference threshold. When the reference threshold is relatively strict (e.g., the lower quartile of the expression distribution of the reference population, i.e., the 25th percentile), the samples judged as low expression are more concentrated in samples with high degree of functional inactivation, resulting in higher specificity of the included response candidates, but a corresponding decrease in sensitivity; when the reference threshold is relatively lenient (e.g., the median of the expression distribution of the reference population, i.e., the 50th percentile), the range of samples judged as low expression is wider, resulting in higher sensitivity, but some samples with lower degree of functional inactivation may be included. Those skilled in the art can select the reference threshold within the above quantile range according to the specific application's requirements for sensitivity and specificity. Given that the PoC data show a significant correlation between SMARCA4 expression level and SMARCA2 dependence within the variant sample subset (ρ=0.446, p=4.8e-8), the low expression criterion can indeed enrich samples dependent on SMARCA2. Therefore, under any of the above quantile values, using expression levels below the reference threshold as the criterion for response candidates is valid, with the only difference being the trade-off between sensitivity and specificity.

[0066] Statistical Methods: The significance of differences in SMARCA2 dependence among sample groups in this PoC was assessed using the Mann-Whitney nonparametric test (also known as the Wilcoxon rank-sum test). This test does not rely on the assumption of normality of data and is suitable for situations where sample sizes are unbalanced in this study (1052 wild-type cases, 80 missense cases, and 60 truncated loss-of-function cases). The association between SMARCA4 expression levels and SMARCA2 dependence within the variant sample subset was assessed using the Spearman rank correlation coefficient, a method robust to variable distribution patterns and suitable for characterizing monotonic relationships. The proportion of strong dependence was defined as the percentage of samples in each group with a SMARCA2 CRISPR gene effect less than -0.5. These statistics and significance levels collectively support the feasibility of screening SMARCA2-dependent responders based on variant functional consequence categories and expression levels.

[0067] For ease of comparison, the key data of this embodiment are summarized in the following table:

[0068] Wild type (WT) 1052 1.4% — Class II missense 80 7.5% p=0.071 (not significant) Truncated loss of function (Class I LoF) 60 33.3% p = 7.8e-9 (significant)

[0069] Note: The truncated loss-of-function category showed a significant difference relative to the missense category, p=2.1e-4; within the variant subset (n=137), SMARCA4 expression levels were correlated with SMARCA2 dependence (Spearman ρ=0.446, p=4.8e-8). Strong SMARCA2 dependence was defined as CRISPR gene effect < -0.5.

[0070] Example 3: Implementation of the Detection Kit and Data Processing System

[0071] This embodiment provides an implementation of the detection kit and the in vitro data processing system for non-diagnostic and non-therapeutic purposes of the present invention. As shown in Figure 4, the kit and system include modules M1 to M8.

[0072] Implementation of the detection kit. The detection kit of this embodiment is used to screen response candidates for selective degradation agents or inhibitors of SMARCA2. It includes reagents for determining the functional consequence category of SMARCA4 gene mutations and reagents for measuring the expression level of SMARCA4 expression products. The detection kit is used to implement the method described in Example 1.

[0073] The reagents used to determine the functional consequences of SMARCA4 gene variants are nucleic acid sequencing reagents, which may include capture probes or amplification primers targeting the SMARCA4 gene region, library construction reagents, sequencing reaction reagents, etc. In conjunction with the variant effect annotation process, the detection and functional consequences classification of SMARCA4 variants in samples can be achieved (distinguishing between at least truncated loss-of-function categories and missense categories).

[0074] Reagents used to determine the expression level of SMARCA4 expression products include RNA quantification reagents or SMARCA4 protein immunohistochemical reagents. The former may include specific primers and probes for SMARCA4 mRNA, reverse transcription and amplification reagents, and corresponding internal control reagents; the latter may include anti-SMARCA4 (BRG1) antibody and corresponding immunohistochemical colorimetric and control reagents.

[0075] The detection kit may also include a control or internal reference for determining a reference threshold, such as a control sample from a SMARCA4 functionally inactivated reference population, a standard with known expression levels, or an internal reference gene detection reagent for expression normalization, so as to establish and calibrate a reference threshold for low expression determination in actual detection.

[0076] Data processing system implementation. The data processing system of this embodiment is used to implement the data processing flow involved in steps S2 to S8 of Embodiment 1, and the in vitro non-diagnostic and non-therapeutic data processing method described in the fourth aspect of the present invention. As shown in FIG4, the system includes the following modules.

[0077] The sample receiving and nucleic acid extraction module M1 is configured to receive in vitro biological samples and prepare nucleic acid or slice objects for mutation detection and expression detection, corresponding to step S1.

[0078] The variant detection and functional effect annotation module M2 is configured to process the sequencing data of the sample nucleic acid, detect the SMARCA4 variant, and annotate its functional effect, corresponding to step S2.

[0079] The variant functional consequence classification module M3 is configured to classify SMARCA4 variants into either the truncated loss of function category or the missense category based on the annotation results, corresponding to step S3.

[0080] The expression level determination module M4 is configured to acquire and process SMARCA4 expression product expression level data, corresponding to step S4.

[0081] The expression level and reference threshold comparison module M5 is configured to compare the measured expression level with the reference threshold and determine whether it is lower than the reference threshold, corresponding to step S5.

[0082] The functional inactivation degree comprehensive determination module M6 is configured to comprehensively compare the functional consequences of mutations with the expression level results, determine the degree of SMARCA4 functional inactivation, and identify response candidates or output molecular typing identifiers, corresponding to step S6.

[0083] The SMARCA2 function dependency scoring module M7 is configured to optionally calculate the SMARCA2 function dependency score by weighting the variant functional consequence category and the normalized expression level and include it in the judgment, corresponding to step S7.

[0084] The result output module M8 is configured to output response candidate determination results or molecular subtyping identifiers, corresponding to step S8. In the implementation method of data processing, the output of the result output module M8 is limited to a molecular subtyping identifier characterizing the degree of functional inactivation of SMARCA4 in the sample (i.e., whether it belongs to the truncated loss-of-function category and whether the expression level is lower than the reference threshold), and does not output disease diagnosis conclusions, nor does it include drug administration or treatment steps, thereby complying with the limitation on the output of the data processing method in the fourth aspect of the present invention.

[0085] Each of the above modules can be implemented by a computer program and deployed on the data processing unit of a general computing device or testing instrument. The data is transferred and processed sequentially from module M1 to module M8 according to the data flow shown in Figure 4. Module M7 is an optional module and can be omitted in the implementation where the SMARCA2 functional dependency score is not calculated. The corresponding data is directly transferred from module M6 to module M8.

[0086] Alternative solutions and boundary descriptions

[0087] Without departing from the concept of the present invention, the embodiments of the present invention can be replaced and extended as follows.

[0088] Alternatives to the detection platform. The determination of the functional consequence category of the variant is not limited to the aforementioned targeted next-generation sequencing; any nucleic acid sequencing method capable of detecting SMARCA4 variants and supporting functional effect annotation, such as whole-exome sequencing or multi-gene panel sequencing, can also be used. The determination of expression levels is not limited to the aforementioned RNA quantification or immunohistochemistry; any detection method capable of quantifying or semi-quantitatively analyzing SMARCA4 expression products and comparing them with reference thresholds can be used. The difference between this invention and existing technologies lies in the logic of screening responders based on the degree of SMARCA4 functional inactivation (variant functional consequence category + expression level successive proxy), rather than a specific detection platform; changing the detection platform does not alter the essence of this invention.

[0089] Tumor type alternatives. The subjects of this invention are not limited to non-small cell lung cancer and esophageal cancer. Any solid tumor in which SMARCA4 function loss can occur can be screened for SMARCA2 selective degrader or inhibitor response candidates using this method.

[0090] Extending the judgment logic. In addition to the basic logic that a candidate is included if either the truncated loss of function category or the expression level is below the reference threshold, the SMARCA2 functional dependency score can be superimposed for continuous stratification as described in step S7; or more detailed subcategories can be included in the functional consequence category (such as further distinguishing the location of truncation and the scope of splicing influence), as long as it falls within the dichotomy framework of the truncated loss of function category and the missense category, it falls within the scope of this invention.

Claims

1. A method for detecting the degree of functional inactivation of SMARCA4 in an in vitro sample, used to screen response candidates for selective degradation agents or inhibitors of SMARCA2 from test subjects, characterized in that, include: For in vitro biological samples from the subjects, the functional consequence categories of SMARCA4 gene variants are determined, which include at least truncated loss-of-function categories and missense categories, wherein the truncated loss-of-function categories include nonsense mutations, frameshift mutations and splice site variants, and the missense categories include missense mutations and whole-frame mutations; The expression level of SMARCA4 expression product in the ex vivo biological sample was detected, and the expression level was compared with a reference threshold. The degree of SMARCA4 functional inactivation is determined based on the functional consequence category and the expression level. When the SMARCA4 gene mutation belongs to the truncated loss of function category and / or the expression level is lower than the reference threshold, the test subject is identified as a candidate for response to a selective SMARCA2 degrader or inhibitor. The SMARCA2 selective degrader or inhibitor is any degrader or inhibitor that is selective for SMARCA2, and is not limited to a specific compound.

2. The method according to claim 1, characterized in that, The truncated loss of function category is a variant annotated as HIGH impact or LikelyLoF based on the variant effect prediction; the missense category is determined to be a possible function retention, and when it exists alone without the expression level being lower than the reference threshold, it is not included in the criteria for identifying the subject as a response candidate.

3. The method according to claim 1, characterized in that, The expression level of the SMARCA4 expression product is the expression level of SMARCA4 mRNA or the expression level of SMARCA4 protein; the reference threshold is a threshold set relative to the expression level distribution of a reference population of SMARCA4 that is not functionally inactivated, and the expression level below the reference threshold indicates an enhanced degree of functional inactivation of SMARCA4.

4. The method according to claim 1, characterized in that, The determination of the functional consequence category of the SMARCA4 gene variant is achieved by performing nucleic acid sequencing on the in vitro biological sample and annotating the functional effect of the detected variant; the detection of the expression level of the SMARCA4 expression product is achieved by RNA quantification or immunohistochemistry.

5. The method according to claim 1, characterized in that, The subjects were individuals with non-small cell lung cancer or esophageal cancer.

6. The method according to claim 1, characterized in that, Also includes: The SMARCA2 functional dependency score is determined based on the functional consequence category and the expression level, and the SMARCA2 functional dependency score, together with the functional consequence category and the expression level, is used to determine the response candidates. The SMARCA2 functional dependency score is a continuous value that characterizes the degree of dependence of the sample on SMARCA2.

7. The method according to claim 1, characterized in that, The SMARCA2 selective degrader is a selective SMARCA2 / BRM protein degradation targeting chimera (PROTAC) type degrader.

8. A detection kit for screening candidates for selective degradation agents or inhibitors of SMARCA2, characterized in that, include: A reagent for determining the functional consequence category of SMARCA4 gene variants, wherein the functional consequence category includes at least a truncated loss of function category and a missense category; And reagents for determining the expression level of SMARCA4 expression products; the detection kit is used to implement the method of any one of claims 1 to 7.

9. The detection kit according to claim 8, characterized in that, The reagent used to determine the functional consequence category of SMARCA4 gene mutation is a nucleic acid sequencing reagent; the reagent used to determine the expression level of SMARCA4 expression product is an RNA quantitative reagent or a SMARCA4 protein immunohistochemical reagent; the detection kit also includes a control or internal control for determining the reference threshold.

10. The use of a combination of reagents for determining the functional consequence category of SMARCA4 gene variants and reagents for measuring the expression level of SMARCA4 expression products in the preparation of a detection kit for screening candidates for selective degradation agents or inhibitors of SMARCA2, characterized in that, The functional consequence categories include at least a truncated loss-of-function category and a missense category. The truncated loss-of-function category includes nonsense mutations, frameshift mutations, and splice site variations. The detection kit is configured to indicate that the corresponding test subject is a candidate response to a selective degrader or inhibitor of SMARCA2 when the SMARCA4 gene variation belongs to the truncated loss-of-function category and / or the expression level of the SMARCA4 expression product is below a reference threshold.

11. A data processing method for in vitro, non-diagnostic, and non-therapeutic purposes, characterized in that, include: The method involves acquiring SMARCA4 gene variant annotation data and SMARCA4 expression product expression level data for a sample; determining whether the SMARCA4 variant in the sample belongs to the truncated loss-of-function category based on the variant annotation data, where the truncated loss-of-function category includes nonsense mutations, frameshift mutations, and splice site mutations, or variants annotated as HIGH impact or LikelyLoF based on variant effect prediction; comparing the expression level data with a reference threshold; and outputting a molecular typing identifier based on the above results, whereby the molecular typing identifier characterizes the degree of SMARCA4 functional inactivation in the sample, including whether it belongs to the truncated loss-of-function category and whether the expression level is below the reference threshold; the output of the data processing method is limited to the molecular typing identifier, and does not output disease diagnosis conclusions or include drug administration or treatment steps.

Images