Method for constructing sensitivity prediction model for neoadjuvant chemotherapy of advanced pharyngolaryngeal cancer

Abstract

Provided in the present invention is a method for constructing a sensitivity prediction model for neoadjuvant chemotherapy of advanced pharyngolaryngeal cancer. A prediction model is constructed by means of identifying genes of patients with advanced pharyngolaryngeal cancer before neoadjuvant chemotherapy, which genes are differentially expressed in both a sensitive group and a resistant group and in cancerous tissues and adjacent normal tissues. The model can predict, before treatment, the sensitivity of patients to neoadjuvant chemotherapy, so as to identify chemotherapy responders, and can design individualized treatment plans for the patients, so as to perform precise treatment, thus retaining the pharyngolaryngeal function of the patients to the greatest extent while ensuring or improving the survival rate, thereby greatly improving the quality of life of the patients. At present, the prediction model is undergoing a multi-center prospective study, which helps to promote the model more widely and realize earlier clinical translation.

Description

A method for constructing a prediction model for neoadjuvant chemotherapy sensitivity of middle and advanced stage laryngopharyngeal cancer TECHNICAL FIELD

[0001] The present application relates to the technical field of biological medicine, in particular to a method for constructing a prediction model for neoadjuvant chemotherapy sensitivity of middle and advanced stage laryngopharyngeal cancer. BACKGROUND

[0002] Laryngopharyngeal cancer is a common and frequently occurring squamous cell carcinoma of the head and neck, and its incidence is gradually increasing worldwide. Under the existing comprehensive treatment system for tumors, the risk of local recurrence and distant metastasis in patients with middle and advanced stage laryngopharyngeal cancer is still as high as 15-40%, and the 5-year overall survival rate is only about 25-35%. According to the clinical practice guidelines of the National Comprehensive Cancer Network (NCCN), the treatment plan for patients with middle and advanced stage laryngopharyngeal cancer is not uniform, and the guidelines recommend surgical treatment, induction chemotherapy, concurrent chemoradiotherapy, and comprehensive treatment, which almost include all tumor treatment options. Surgical treatment can completely remove the tumor, but it also causes the patient to lose the ability to speak and swallow, resulting in disability for the patient, causing great physical and mental trauma to the patient, and also causing a heavy medical and economic burden to society; non-surgical treatment options, such as radiotherapy or concurrent chemoradiotherapy, and salvage total laryngectomy after treatment failure, can lead to serious complications. In recent years, neoadjuvant chemotherapy has been applied to clinical practice, and the next step of treatment is selected based on the efficacy of neoadjuvant chemotherapy, i.e., radiotherapy or concurrent chemoradiotherapy for patients sensitive to neoadjuvant chemotherapy, and surgical treatment for patients insensitive to neoadjuvant chemotherapy. Compared with the option of surgical treatment, this option can significantly increase the chance of preserving laryngopharyngeal function without reducing survival rate.

[0003] Neoadjuvant chemotherapy refers to the use of adjuvant chemotherapy before local treatment, i.e., surgery or radiotherapy, which can reduce tumor burden in a short period of time. However, a large number of clinical observations have shown that some patients are not sensitive to neoadjuvant chemotherapy, and they experience the pain of chemotherapy without benefiting from it. Why are some patients sensitive to chemotherapy and some patients not? This may be related to the heterogeneity of tumors among individuals. The same type of pathology at the same site may not have exactly the same microenvironment and dominant signal pathway for cell proliferation regulation in different individuals. In recent years, based on the heterogeneity of tumors, researchers of head and neck tumors have reported that some molecular markers can be used to predict the efficacy of neoadjuvant chemotherapy and the prognosis of patients. However, in the diagnosis and treatment of laryngopharyngeal cancer, there is still a lack of a high-precision diagnostic prediction model for the sensitivity of neoadjuvant chemotherapy in patients with middle and advanced stage laryngopharyngeal cancer. SUMMARY

[0004] The present application aims to provide a method for constructing a prediction model for the sensitivity of neoadjuvant chemotherapy of middle and advanced stage laryngopharyngeal cancer, so as to fill the gap in the prediction technology of the sensitivity of neoadjuvant chemotherapy of middle and advanced stage laryngopharyngeal cancer, provide individualized treatment plan for patients, and ensure or improve the survival rate and quality of life of patients.

[0005] To achieve the above technical purposes, the technical scheme adopted by the present application is as follows:

[0006] In a first aspect, the present application provides a method for constructing a prediction model for the sensitivity of neoadjuvant chemotherapy of middle and advanced stage laryngopharyngeal cancer, comprising the following steps:

[0007] Step 1: Collecting cancer tissues and para-cancer normal tissues of patients with middle and advanced stage laryngopharyngeal cancer before neoadjuvant chemotherapy, and performing RNA-seq sequencing analysis to obtain the differentially expressed genes of the two;

[0008] Step 2: Obtaining the sensitivity information of the patients with middle and advanced stage laryngopharyngeal cancer after neoadjuvant chemotherapy in step 1, and dividing the aforementioned patients into a sensitive group and a resistant group according to the sensitivity information, and performing RNA-seq sequencing analysis on the cancer tissues of the sensitive group and the resistant group to obtain the differentially expressed genes of the two;

[0009] Step 3: Comparing the differentially expressed genes obtained in steps 1 and 2 and obtaining the differential genes with intersection;

[0010] Step 4: Using a random forest model to sort the importance of the differential genes with intersection obtained in step 3, and drawing an AUC curve to screen out genes with AUC>0.8;

[0011] Step 5: Further verifying the relative expression of the genes obtained in step 4 by qPCR to screen out genes related to the sensitivity of neoadjuvant chemotherapy of middle and advanced stage laryngopharyngeal cancer, and constructing a prediction model.

[0012] Preferably, in step 1, the differentially expressed genes of the cancer tissues and the para-cancer normal tissues have |fold change|>2.0 and P<0.05.

[0013] Preferably, in step 2, the differentially expressed genes of the sensitive group and the resistant group have |fold change|>2.0 and P<0.05.

[0014] Preferably, in step 5, the genes related to the neoadjuvant chemotherapy of middle and advanced stage laryngopharyngeal cancer include NRIP1, GIMAP7, CD72, THBS4, ABCA9 and SNED1.

[0015] Preferably, the prediction model is based on a risk score formula for result prediction, the risk score formula is: score=0.940*ABCA9+0.800*CD72+0.917*GIMAP7+0.934*NRIP1+0.765*SNED1+0.810*THBS4, Cutoff=0.0888, Score<0.0888 is sensitive, and Score> / =0.0888 is insensitive.

[0016] In a second aspect, the present application provides the prediction model of neoadjuvant chemotherapy sensitivity of middle-late stage laryngopharyngeal cancer obtained by the construction method.

[0017] In a third aspect, the present application provides a biomarker for predicting the neoadjuvant chemotherapy sensitivity of middle-late stage laryngopharyngeal cancer, the biomarker comprising: NRIP1, GIMAP7, CD72, THBS4, ABCA9 and SNED1.

[0018] In a fourth aspect, the present application provides the use of the biomarker in the third aspect in constructing the prediction model in the second aspect.

[0019] In a fifth aspect, the present application provides a kit for evaluating the neoadjuvant chemotherapy sensitivity of middle-late stage laryngopharyngeal cancer, comprising the prediction model of neoadjuvant chemotherapy sensitivity of middle-late stage laryngopharyngeal cancer in the second aspect or the biomarker in the third aspect.

[0020] Preferably, when the kit comprises the prediction model of neoadjuvant chemotherapy sensitivity of middle-late stage laryngopharyngeal cancer, the prediction model is based on a risk score formula for result prediction, the risk score formula is: score=0.940*ABCA9+0.800*CD72+0.917*GIMAP7+0.934*NRIP1+0.765*SNED1+0.810*THBS4, Cutoff=0.0888, Score<0.0888 is sensitive, and Score> / =0.0888 is insensitive; when the kit comprises the biomarker, it further comprises reagents for detecting the expression of the biomarker.

[0021] Compared with the prior art, the present application has the following beneficial effects:

[0022] The application provides a method for constructing a prediction model for neoadjuvant chemotherapy sensitivity of middle and advanced stage laryngeal cancer, which is constructed by identifying genes differentially expressed in both the sensitive group and the resistant group and in the cancer and the normal tissue adjacent to the cancer of the middle and advanced stage laryngeal cancer patients before neoadjuvant chemotherapy, and the model can predict the neoadjuvant chemotherapy sensitivity of the patients before treatment, and further identify the chemotherapy beneficiaries, so that the individualized treatment plan can be designed for the patients, and the laryngeal function of the patients can be preserved as much as possible under the premise of ensuring or improving the survival rate, and the life quality of the patients is greatly improved. At present, the prediction model is in a multicenter prospective study, which is helpful to the wider promotion of the model and the earlier realization of clinical transformation. BRIEF DESCRIPTION OF DRAWINGS

[0023] FIG. 1 is the RNA-seq sequencing analysis result of the cancer tissue and the tissue adjacent to the cancer in the embodiment 1 of the application, wherein A is the volcano plot of the difference analysis of the cancer and the normal tissue adjacent to the cancer, B is the volcano plot of the difference analysis of the cancer tissue in the sensitive group and the resistant group, C is the intersection of the two groups of differentially expressed genes in A and B, and D is the expression heat map of the intersection genes, which are differentially expressed in both the sensitive group and the resistant group of the neoadjuvant chemotherapy and in the cancer and the normal tissue adjacent to the cancer.

[0024] FIG. 2 is the 10 genes with the area under the curve (AUC) greater than 0.8 in the embodiment 1 of the application.

[0025] FIG. 3 is the verification of the relative expression of the 10 genes with the area under the curve (AUC) greater than 0.8 screened out in the embodiment 1 of the application by qPCR in another independent group (55 cases in the sensitive group and 57 cases in the resistant group), and 6 genes (NRIP1, GIMAP7, CD72, THBS4, ABCA9 and SNED1) are screened out, which are significantly highly expressed in the insensitive tissue compared with the sensitive group, indicating that the expression of the 6 genes can predict the sensitivity of the patients to the neoadjuvant chemotherapy.

[0026] FIG. 4 is the performance test of the 6-gene prediction model in the embodiment 1 of the application, wherein A is the comparison of the area under the curve (AUC) of the 6-gene prediction model and the common clinical indicators, B is the expression heat map of the 6 genes, the patients can be divided into high and low expression groups according to the expression of the six genes by K-means clustering, and C is the survival analysis of the patients in the high and low expression groups. DETAILED DESCRIPTION

[0027] The application will be further described in detail in combination with the drawings and specific embodiments, which are an explanation but not a limitation of the application.

[0028] Embodiment 1

[0029] The embodiment provides a method for constructing a prediction model for neoadjuvant chemotherapy sensitivity of middle-late stage laryngopharyngeal cancer, and specifically comprises the following steps.

[0030] (1) Collect 54 cases of middle-late stage laryngopharyngeal cancer patients before neoadjuvant chemotherapy (22 cases of sensitive group, 32 cases of resistant group, all from Beijing Tongren Hospital Affiliated to Capital Medical University) tissue specimens (cancer tissue and paracancer tissue), and perform RNA-seq sequencing analysis, screen the differentially expressed genes (|fold change|>2.0, P<0.05), and the results are shown in FIG. 1, 751 differential genes are obtained (FIG. 1, A), 636 differential genes are obtained (FIG. 1, B), and 107 differential genes are obtained (FIG. 1, C) in the intersection of the two groups, which are differentially expressed in the sensitive group and the resistant group, and differentially expressed in the cancer and the paracancer tissue (FIG. 1, D).

[0031] (2) The random forest model (constructed by the random forest regression method of the R package randomForest in R language, which is a conventional construction method in the art, and will not be described in detail here) is used to sort the 107 genes in the order of importance from large to small, and 10 genes with an area under the curve (AUC) greater than 0.8 are screened out (as shown in FIG. 2). Further collect 112 cases of middle-late stage laryngopharyngeal cancer patients before neoadjuvant chemotherapy (55 cases of sensitive group, 57 cases of resistant group, all from Beijing Tongren Hospital Affiliated to Capital Medical University) cancer tissue as an external validation set (the traditional clinical indicators of the patients are shown in the table of FIG. 4, A), verify the relative expression of the 10 genes in the external samples by qPCR, and screen 6 genes (NRIP1, GIMAP7, CD72, THBS4, ABCA9 and SNED1). Compared with the sensitive group, the expression of the 6 genes in the insensitive tissue is significantly high, indicating that the expression of the 6 genes can predict the sensitivity of the patient to neoadjuvant chemotherapy (as shown in FIG. 3).

[0032] In step (2), the method of qPCR is as follows:

[0033] After collecting the cancer tissue, RNA extraction is performed, the kit is Nuoyuan R711, and the RNA is extracted according to the instruction manual, then a reverse transcription reaction is performed to obtain cDNA, and the kit is Nuoyuan R323; the product is immediately used for qPCR reaction, and the primer sequences, reaction conditions and system are shown in Tables 1-3, and 2xTaq Pro Universal SYBR qPCR Master Mix is Nuoyuan Q712 product.

[0034] Table 1 qPCR primer sequence

[0035] In Table 1, GAPDH is an internal reference gene.

[0036] Table 2 qPCR reaction conditions

[0037] Table 3 qPCR reaction system

[0038] (3) As shown in Figure 4, a prediction model was constructed by screening the six genes, and the area under the curve reached 0.949, which was significantly better than the traditional clinical indicators (Figure 4, panel A); according to the expression of the six genes, the patients could be divided into high and low expression groups by K-means clustering (Figure 4, panel B), and the overall survival of the patients in the high expression group was significantly worse than that in the low expression group (Figure 4, panel C). This shows that the 6-gene prediction model not only can accurately predict the efficacy of neoadjuvant chemotherapy for patients with advanced laryngopharyngeal cancer, but also can early identify patients with poor prognosis, assist doctors in making clinical decisions, adjust the neoadjuvant chemotherapy regimen, determine the timing of surgery, and promote the precision of laryngopharyngeal cancer treatment. It is expected to provide more personalized surgery and nursing stratified management for patients, and reduce the side effects and economic burden of chemotherapy.

[0039] Example 2

[0040] Prospective multicenter clinical study: Patients with T2-4a advanced laryngopharyngeal cancer undergoing neoadjuvant chemotherapy were enrolled, and pathological biopsy tissue specimens were used for neoadjuvant chemotherapy sensitivity detection. According to the detection results, the patients were divided into sensitive and resistant groups, and the actual neoadjuvant chemotherapy efficacy of the patients was compared. At present, a total of 136 patients have been enrolled, and the data are shown in Table 4.

[0041] According to the 6-gene prediction model obtained above, the score of the patient was calculated score = 0.940 x ABCA9 + 0.800 x CD72 + 0.917 x GIMAP7 + 0.934 x NRIP1 + 0.765 x SNED1 + 0.810 x THBS4, Cutoff = 0.0888, Score < 0.0888 is sensitive, and Score> / = 0.0888 is not sensitive. The expression of the six genes in the newly collected laryngopharyngeal cancer tissue specimens before neoadjuvant chemotherapy was detected by qPCR after reverse transcription of RNA, and the ACq value of the six genes (internal reference gene is GAPDH) was calculated. The score was calculated by substituting into the formula, and the sensitivity was judged by checking its value in the experience distribution of sensitivity and non-sensitivity. Among them, the sensitivity = true positive number / (true positive number + false negative number) x 100%. The specificity = true negative number / (true negative number + false positive number) x 100%.

[0042] Table 4 Neoadjuvant chemotherapy sensitivity results

[0043] In Table 4, the 6-gene prediction model obtained from Example 1 was used to predict the sensitivity of 136 patients to neoadjuvant chemotherapy, of which 62 patients were sensitive to neoadjuvant chemotherapy and 74 patients were resistant to neoadjuvant chemotherapy. The 136 patients all underwent actual neoadjuvant chemotherapy, and the efficacy results showed that of the 62 patients predicted to be sensitive by the model, 51 were actually sensitive and 11 were resistant, and of the 74 patients predicted to be resistant by the model, 60 were actually resistant and 14 were sensitive. The sensitivity = 51 / 65 = 78.5%, the specificity = 60 / 71 = 84.5%, the Youden's index = sensitivity + specificity - 1 = 0.63, indicating that the model has good prediction performance.

[0044] The present application also provides a late-stage laryngeal cancer induction chemotherapy sensitivity evaluation kit, comprising the 6-gene prediction model obtained from Example 1, and the results are predicted based on the risk score formula: score = 0.940*ABCA9 + 0.800*CD72 + 0.917*GIMAP7 + 0.934*NRIP1 + 0.765*SNED1 + 0.810*THBS4, Cutoff = 0.0888, Score < 0.0888 is sensitive, and Score> / = 0.0888 is not sensitive. The kit also covers primers and reagents for detecting the expression of the 6-gene model in tissues. The kit can help patients predict the sensitivity of induction chemotherapy, thereby identifying chemotherapy beneficiaries, and can design individualized treatment plans for patients and provide precise treatment.

[0045] The above-described examples only express several embodiments of the present application, which are described in more detail and in more detail, but should not be construed as limiting the scope of the present patent. It should be noted that for ordinary skilled persons in the art, without departing from the concept of the present application, several modifications and improvements can be made, which are within the scope of the present application. Therefore, the protection scope of the present patent should be subject to the appended claims.

Claims

1. A method for constructing a predictive model for the sensitivity of neoadjuvant chemotherapy in intermediate and advanced laryngeal cancer, characterized in that, Includes the following steps: Step 1: Collect cancer tissue and adjacent normal tissue from patients with intermediate and advanced laryngeal cancer before neoadjuvant chemotherapy, and perform RNA-seq sequencing analysis to obtain differentially expressed genes between the two. Step 2: Obtain the sensitivity information of patients with intermediate and advanced laryngeal cancer after neoadjuvant chemotherapy in Step 1, and divide the aforementioned patients into a sensitive group and a tolerant group based on the sensitivity information. Perform RNA-seq sequencing analysis on the cancer tissues of the sensitive group and the tolerant group to obtain the differentially expressed genes between the two groups. Step 3: Compare the differentially expressed genes obtained in Step 1 and Step 2 and obtain the differentially expressed genes with overlap; Step 4: Use a random forest model to rank the differentially expressed genes with intersection obtained in Step 3 by importance, and plot the AUC curve to screen out genes with AUC > 0.

8. Step 5: The genes obtained in Step 4 are further validated for relative expression by qPCR, and genes related to the sensitivity of neoadjuvant chemotherapy in intermediate and advanced laryngeal cancer are screened to construct a predictive model.

2. The construction method according to claim 1, characterized in that, In step 1, the differentially expressed gene fold change between cancerous tissue and adjacent normal tissue was >2.0, P<0.

05.

3. The construction method according to claim 1 or 2, characterized in that, In step 2, the differentially expressed gene fold change between the sensitive group and the tolerant group was >2.0, P<0.

05.

4. The construction method according to claim 1, characterized in that, In step 5, the genes associated with neoadjuvant chemotherapy for intermediate and advanced laryngeal cancer include: NRIP1, GIMAP7, CD72, THBS4, ABCA9, and SNED1.

5. The construction method according to claim 4, characterized in that, The prediction model predicts results based on a risk scoring formula, which is: score = 0.940 × ABCA9 + 0.800 × CD72 + 0.917 × GIMAP7 + 0.934 × NRIP1 + 0.765 × SNED1 + 0.810 × THBS4, Cutoff = 0.0888, where Score < 0.0888 indicates sensitivity, and Score > / = 0.0888 indicates insensitivity.

6. The neoadjuvant chemotherapy sensitivity prediction model for intermediate-to-advanced laryngeal cancer obtained by the construction method according to any one of claims 1 to 5.

7. A biomarker for predicting the sensitivity of neoadjuvant chemotherapy in intermediate-to-advanced laryngeal cancer, characterized in that, The biomarkers include: NRIP1, GIMAP7, CD72, THBS4, ABCA9, and SNED1.

8. The application of the biomarker of claim 7 to the predictive model of claim 6.

9. A kit for assessing the sensitivity of neoadjuvant chemotherapy in intermediate and advanced laryngeal cancer, characterized in that, This includes the neoadjuvant chemotherapy sensitivity prediction model for intermediate and advanced laryngeal cancer as described in claim 6 or the biomarker as described in claim 7.

10. The reagent kit according to claim 9, characterized in that, When the kit includes the neoadjuvant chemotherapy sensitivity prediction model for intermediate and advanced laryngeal cancer as described in claim 6, the prediction model predicts the outcome based on a risk score formula, which is: score = 0.940 × ABCA9 + 0.800 × CD72 + 0.917 × GIMAP7 + 0.934 × NRIP1 + 0.765 × SNED1 + 0.810 × THBS4, Cutoff = 0.0888, Score < 0.0888 indicates sensitivity, and Score > 0.0888 indicates insensitivity; when the kit includes the biomarker as described in claim 7, it also includes reagents for detecting the expression of the biomarker.

Images