Individualized Anti-tumor Treatment Of Angiogenesis Targeted Therapies Based On Colorectal Cancer PDX Models

Background:Colorectal cancer(CRC)is the fourth cause of cancer-related deaths worldwide,and the targeted and personalized therapies have taken CRC into a new age of treatment.However,in fact,not all patients benefit from current therapies,thus more personalized medicine appoaches remain to be developed,therefore there is an urgent need for discovering more effective predictive biomarkers and targets for more individualized precision medicine.Patient-derived xenograft(PDX)models,have been increasingly widely used in various types of cancers for translational research in recent years.PDXs are established by directly transplanting fresh patient tumor tissue into immune-compromised mice.Accumulating evidences indicate that PDX models can maintain major pathological and molecular characteristic of the original tumor,thus to be an important tool for cancer precision medicine and translational research,including drug screening,personalized medicine application,biomarker development,understanding of mechanisms of drug resistance and prospectively identifcation of clinical translation hypotheses.Omics technologies can be a powerful tool to understand the genomic landscape of patients,discover drug predictive biomarkers and investigate mechanisms of drug resistance.Integration of multiple omics technologies,which can compensate for the missing or noise in any single high-throughput data type,will increase the success rate of effective biomarker discovery for targeted therapies.Over the past few years,several clinical trials have demonstrated the clinical benefits of anti-angiogenic agents on cancers.Vascular endothelial growth factor-A(VEGF-A,usually referred to as VEGF),a key proangiogenic growth factor,is known to play an important role in angiogenic processes through both direct or indirect mechanisms.VEGF inhibitors have been validated effective at inhibiting angiogenesis in many types of cancers.However,despite the clinical success of anti-VEGF therapies in cancer,primay or secondary resistance to anti-VEGF therapies has become a major challenge for anti-angiogenic therapies.Multiple resistance mechanisms cause insensitivity of tumor vasculature to anti-VEGF therapies and/or decreased dependence of tumors on angiogenesis.There is an urgent need to discover predictive biomarkers for anti-VEGF therapies and further understand the mechanisms that mediate resistance to anti-angiogenic therapies.In addition,the success of combined therapies of anti-VEGF therapies and chemotherapies relies on the "normalization" of architecture and function of existing vasculature.The normalization window,defined as a period of time when blood flow and oxygen-supplying transitorily increases,is dose-and time-dependent.Therefore,more studies regarding biomarkers of vascular normalization window will be critical for optimizing the efficacy of combined anti-VEGF and cytotoxic therapies.A Novel VEGF-trap FP3(also named as KH903)is an engineered protein which contains the extracellular domain 2 of VEGF receptor 1(Flt-1)and extracellular domain 3 and 4 of VEGF receptor 2(KDR)fused to the Fc portion of human immunoglobulin G1.In the previous studies,our research group have showed that FP3 has an anti-tumor effect in PDX models of gastric carcinoma and colorectal cancer in nude mice.The aim of this study was to discover and validate efficacy predictive biomarkers for anti-angiogenic targeted therapies as a monotherapy or a combination therapy,thus to provide effective suggestions for individualized treatment.Section IPurpose:The aim of this part was to establish patient-derived colorectal cancer xenograft models for translational research.Methods:Tumor tissues from 85 CRC patients were transplanted into BALB/c nude mice for establishing PDX models.Hematein Eosin staining,immunohistochemical staining(CEA、CK7、ki67、VEGF、FGFR2 和 VEGFR2)and molecular investigations(such as,AmpliSeq Cancer Hotspot Panel Version 2 and NANO-LC/MS/MS)were performed to verify if the PDX retained the main features from the patient.Results:50 out of 85(58.5%)CRC tumors were successfully engrafted.A high genetic concordance between patient donor tumor and PDX was confirmed by Hematein Eosin staining,immunohistochemical staining(CEA、CK7、ki67、VEGF、FGFR2 和 VEGFR2),AmpliSeq Cancer Hotspot Panel Version 2 and NANO-LC/MS/MS.Conclusion:We established 50 CRC PDX models with a high histologic and genetic representativeness of the primary tumors.This platform will represent a reliable tool for CRC precision medicine and cancer translational research.Section ⅡPurpose:The aim of this part was to evaluate a novel anti-vascular agent,FP3,as a single therapy or combined with chemotherapy based on the colorectal cancer PDX models.Methods:A colorectal cancer PDX model was used for the evaluation of FP3.From the 3rd generation,PDX tumors were permitted to grow to a volume of 150-200 mm3,then mice were randomized into 6 groups:1)Saline;2)FP3(20mg/kg iv qw);3)Bevacizumab(20mg/kg iv qw);4)Irinotecan Hydrochloride(5mg/kg ip qw);5)FP3(20mg/kg iv qw)in combination with Irinotecan Hydrochloride(5mg/kg ip qw);6)Bevacizumab(20mg/kg iv qw)in combination with Irinotecan Hydrochloride(CPT-11,5mg/kg ip qw).Then,changes of endothelial cells were investigated by immunohistochemical expressions and immunofluorescence of CD31,and immunohistochemical expressions of ki67 for anti-tumor activity evaluation.Both the normalization effect of FP3 as a single therapy or a combination therapy were evaluated by immunofluorescence of vascular endothelial cells and pericytes.Results:There was a similar RTI(relative tumor inhibition)between the FP3 group and Bevacizumab group,and between the FP3+CPT-11 group and the Bevacizumab+CPT-11 group.While the RTI of combination groups are more significant than single therapy groups.The down-regulation of expression of CD31 and ki67 were found after FP3 treatment.When combined with CPT-11,the down-regulation of expression of ki67 was more significant,but not for CD31.Vascular normalization effect was found in both single therapy group and combination therapy group.Conclusion:FP3 inhibited the angiogenesis and then suppressed tumor growth.When combined with CPT-11,the synergistic effect of anti-tumor growth was significant,while the anti-angiogenesis effect was similar to the the single FP3 therapy.FP3 showed a similar anti-tumor growth effect with Bevacizumab,either as a single or a combined therapy.Vascular normalization was found in both single FP3 group and combination group.Section ⅢPurpose:The aim of this part was to discover predictive biomarkers for FP3 based on colorectal cancer PDX models and high-throughput omics technologies.Methods:The efficacy of FP3 as a high dosage(60mg/kg iv qw)for 20 colorectal cancer PDX models was evaluated and divided into 2 groups(sensitive and non-sensitive).By high-throughput omics technologies,different expressions of molecules as candidate biomarkers were found between two groups.Then validatation of candidate biomarkers were conducted by bioinformatics and prospective analyses.Results:The efficacy of FP3 as a high dosage(60mg/kg iv qw)for 20 colorectal cancer PDX models was evaluated and divided into two groups(sensitive and non-sensitive).Then 3 levels of high-throughput omics technologies(AmpliSeq Cancer Hotspot Panel Version 2,RNA-sequencing and NANO-LC/MS/MS)were used to detect for different expressions of DNA,RNA or proteins as candidate biomarkers.Candidate biomarkers were validatated by bioinformatics and prospective analyses.Finally,HIF-1α protein was found to be the most potential predicive biomarker for FP3.By dividing 20 colorectal cancer PDX models into HIF-1α high expression group and low expression group,the anti-tumor activity of the two groups had a sinificant trend of difference.Then,a HIF-1α inhibitor was evaluated in combination with FP3 for the non-sensitive C70 model(a model with KRAS mutation and high HIF-1α expression),and synergistic effect was sinificant.Immunofluorescence and immunohistochemical expressions shown vascular endothelial cells density and proliferation associated antigen ki67 was downregulated by treatment of FP3 combined with HIF-1α inhibitor.Further,we found by western blot that HIF-1α and VEGF expression was suppressed after combined treatment of HIF-1α inhibitor;for single FP3 treatment,HIF-1α was upregulated,while VEGF was suppressed temporarily but bound several days later.Conclusion:HIF-1α and its assosiated pathway might be potential predictive biomarker of FP3 efficacy.High expression of HIF-1α,which contribute to the VEGF sustainedly overexpression,might be the mechanism of FP3 resistance.HIF-1α inhibitor combined thrapay might rescue the resistance caused by high expression of HIF-1α.Section ⅣPurpose:The aim of this part was to validatate the effectiveness of serum Ang-1/Ang-2 consentration ratio as a predictive biomarker for anti-angiogenesis therapies in combination with chemotherapeutics.Methods:A colorectal cancer PDX which was evaluated as non-sensitive to the single FP3 therapy was used.Dose-escalation study of FP3(7.5,15,30 and 60 mg/kg)combined with CPT-11 was performed to discover the optimal dosage regimen.Serum Ang-1 and Ang-2 expression were detected by ELISA.A potential correlation between drug responses and Ang-1/Ang-2 ratio were analyzed.Results:FP3(15 mg/kg,i.v.qw)combined with CPT-11 was found to be the optimal dosage regimen.All dosages of FP3 groups showed an ascending Angl/Ang2 ratio after drug administration,while ascending velocity were less significant in 7.5 and 15 mg/kg FP3 groups than that in 30 and 60 mg/kg FP3 groups.The Ang-1/Ang-2 ratio were shown somewhat higher in 15 mg/kg FP3 compared with 7.5 mg/kg FP3 group despite of no statistical significance.Conclusion:We initially validatated the serum Ang-1/Ang-2 consentration ratio as the biomarker of vascular normalization for anti-angiogenesis therapies ib combination with chemotherapeutics.