Screening Pathogenic Genes By WES And Investigating The Influence Of RhTPO On Megakaryocytopoiesis And Thrombocytopoiesis In ITP

Part ?:Screening pathogenic genes for ITP and its subgroup by whole exome sequencingBackground:Primary immune thrombocytopenia(ITP)is an acquired hemorrhagic disease whose etiology and pathogenesis are not fully elucidated.The common characteristic of the patients was the platelet count below the lower limit of the normal reference value(<100×109/L).Its clinical manifestations are heterogeneous,different patients may show different degrees of bleeding symptoms,and some patients may not have any bleeding symptoms.In patients with bleeding,the skin petechiae and ecchymosis may be mild,and the mucosal bleeding such as oral cavity,digestive tract,urinary tract and even intracranial hemorrhage may occur in the severe cases.In recent years,with the further research on the pathogenesis,clinical diagnosis and treatment of ITP,and the attention on patients' health related quality of life(HRQoL),fatigue,anxiety and cognition impairment symptoms of ITP patients have gradually attracted attention.The pathogenesis of ITP remains to be elucidated.To date,the most critical cause has been recognized as immune intolerance,resulting in increased platelet destruction or insufficient platelet production.It mainly involves humoral and cellular immunity,and involves a variety of immune cells,including B lymphocytes,T lymphocytes,monocytes or macrophages,dendritic cells,mesenchymal stem cells,myeloid derived suppressive cells,etc.The antiplatelet autoantibodies they produce facilitate platelet phagocytosis by macrophages,essentially in the spleen.Macrophages contribute to the perpetuation of the auto-immune response as the main antigen presenting cell during ITP.CD8+T cells also participate to thrombocytopenia by increasing platelet apoptosis.Further studies indicated that antiplatelet autoantibodies could trigger platelet activation and neuraminidase translocation to the cell membrane,causing platelet desialylation and Fc-independent hepatic platelet clearance via Ashwell-Morell receptors.On the other hand,the abnormality of various immune cells,cytokines and platelet antibodies leads to abnormal proliferation,maturation and apoptosis of megakaryocytes,which further affects platelet generation and ultimately leads to platelet deficiency.With more in-depth and comprehensive studies on the pathogenesis of ITP,the complex and interlocked microscopic differences between ITP patients and normal people have been gradually revealed,involving a variety of molecules,cells,organs and systems.Therefore,it is increasingly becoming a key point for researchers in the field of ITP to search for the most essential differences and explore the origin cause of ITP.With the completion of the Human Genome Project,a more complete definition of bioinformatics--a discipline that includes all aspects of the acquisition,processing,storage,distribution,analysis and interpretation of biological information came into being.The large amount of bioinformatics data also promotes the development and maturity of high-throughput sequencing technology and the cross integration of life science and computer science,which further catalyzes and accelerates the development and enrichment of bioinformatics.High-throughput sequencing technology is not the same as bioinformatics,but high-throughput sequencing technology currently leads the development in the field of biological information.Central Dogma is that DNA is transcribed into RNA and then translated into protein,which is the base of high-throughput sequencing technologies.Genome,transcriptome,transcription regulation and proteome sequencing are the key components of high-throughput sequencing technologies and are also the focus of bioinformatics research.Whole genome sequencing(WGS)and whole exon sequencing(WES)are the main research aspects of high-throughput DNA sequencing.The human genome contains about 3.16 billion deoxyribonucleotide base pairs,of which exons account for 1-2%and contain about 85%of known pathogenic mutations.The UCSC version of the hg19 human reference genome contains 23056 genes with the entrez gene ID.Therefore,the detection of all exons in the genome,i.e.,whole-exome sequencing,is of great significance for screening pathogenic mutations associated with the disease.High-throughput sequencing technology has been successfully applied in many fields and a wide range of medical research fields.Studies include genomic variation of esophageal squamous cell carcinoma,genomic heterogeneity of multiple lung cancer,heterogeneity and clonal evolution of bilateral ovarian cancer,exon sequencing to reveal the evolution and origin of hepatocellular cholangiocarcinoma,and single-cell transcriptome sequencing to trace the formation process of hematopoietic stem cells.Genomic sequencing results have rewritten WHO guidelines for AML in 2016,and gene panel has been applied in clinic for acute leukemia,MDS and other diseases.In terms of ITP,there has been no report on sequencing studies with application value in the diagnosis and treatment of ITP.In this study,genomic characteristics and variation information of ITP patients were explored for the first time,and differential gene screening was conducted with known databases and normal healthy controls,so as to search for pathogenic mutant genes of ITP.Objectives:(1)To explore the basic variation characteristics of genomic sequencing data of ITP patients in this study.(2)To explore the differences of total exons between ITP patients and normal controls,including mutation loci and structural variations,and the harmfulness of gene mutation was analyzed.Common mutant genes of ITP patients were screened to obtain pathogenic candidate genes,and gene enrichment and protein interaction analysis were performed to find the key pathways.(3)To find the difference of total exons between ITP patients with abnormal increased megakaryocytes and those with normal.To find the difference of total exons between rhTPO responsive and non-responsive ITP patients.Methods:(1)ITP patients and healthy controls:Patients with typical ITP were screened strictly according to the inclusion and exclusion criteria,and healthy volunteers of the same age and sex were matched.Peripheral blood was collected from patients and healthy volunteers.(2)Whole exome sequencing(WES):Peripheral blood DNA was extracted and sequenced using Illumina Novaseq 6000(Illumina Inc.,San Diego,CA,USA)sequencing platform after DNA sample detection,database capture and library test.(3)Quality Control of sequencing data:Raw data were carefully filtered to clean reads.The resulting fastq data were submitted to in-house quality control software for removing low quality reads,and then were aligned to the reference human genome(GRCh37/hg19)using the Burrows-Wheeler Aligner(BWA),and duplicate reads were marked using sambamba tools.Finally,the coverage,depth and other statistics were calculated.(4)Bioinformatics analysis:Annotation was performed using ANNOVAR(2017June8).Annotations included minor allele frequencies from public control data sets as well as deleteriousness and conservation scores enabling further filtering and assessment of the likely pathogenicity of variants.Harmful mutation loci or genes related to disease were screened out through three analysis strategies based on mutation harmness,sample condition and gene function and phenotype.1)Screening high quality variants and annotation,including SNVs,InDels and CNVs;2)Further screening based on existing databases and software tools;3)ACMG classify;4)Screening shared mutant genes between ITP;5)Correlation sorting for candidate genes of ITP;6)Enrichment analysis for candidate genes;7)Protein-Protein interaction analysis.(5)Verification of analysis results:The candidate pathogenic genes were randomly selected after sequencing and bioinformatics analysis,and the Sanger sequencing was used to confirm the reliability of sequencing results.Extracting RNA form peripheral blood mononuclear cells and reverse transcription.Sanger sequencing was used to detect the cDNA changes which could reflect the RNA.The RNAs of candidate genes were randomly selected for differential expression detection and analysis.Results:(1)Characteristics of sequencing samples:55 ITP patients and 55 age and gender-matched healthy controls were included,including 34 female and 21 male patients.The median age of ITP patients was 39(range:18-80)years,with a median platelet count of 11(range:1-47)×109/L.The median platelet count of healthy controls was 185(range:161-239)×109/L.(2)Quality control information:The mean value of raw data was 11.6 ± 0.97 G.The proportion of total reads on the reference genome reached 99.9%in all sequencing samples.The coverage rate of the target region was 99.6%(range:99.6-99.9%),and the sequencing depth of the target region was 112.0 ± 9.12 X.(3)Gene variation results of ITP sequencing samples:1)The number of SNVs in genome(shown in mean ± SD):Intronic(70788 ± 4827),exonic(22002 ± 151.6),intergenic(20545 ± 2430),ncRNA intronic(5772± 440.4),3'UTR(4008± 218.9),ncRNA exonic(2815± 87.69),5'UTR(2526 ± 84.82),splicing-10bp(2443 ±32.93),lkb upstream of transcription initiation site(2038 ±169.5),1kb downstream of transcription initiation site(1052 ± 99.3),ncRNA splicing-10bp(99.6±5.934).Coding region:synonymous SNVs(11261±82.8),missense SNVs(10199±88.59),unknown region(458.6±29.72),stopgain(74.07±5.799),stoploss(8.564 ±1.642).2)The number of InDels in genome(shown in mean ± SD):Intronic(10322 ±752.3),intergenic(3026 ± 388.6),ncRNA intronic(818.3 ± 72.54),exonic(613.3± 17.79),3'UTR(573.6 ± 36.86),splicing-10bp(466.7 ±16.93),5'UTR(388.5 ±19.14),1kb upstream of transcription initiation site(344.4±29.53),ncRNA exonic(271.8 ± 14.94),1kb downstream of transcription initiation site(151.4 ± 14.86),ncRNA splicing-10bp(10.47±2.098).Coding region:non-frameshift deletion(198.6±10.03),non-frameshift insertion(183.1±10.49),unknown region(90.62±3.325),frameshift deletion(76.56±6.215),frameshift insertion(58.16±4.803),stopgain(5.418±1.512),stoploss(0.818±0.722).3)The number of CNVs in genome(shown in mean±SD):Duplicated CNVs(9±4.31),coding sequence(CDS)(7±3.34);deleted CNVs(4±2.63),coding sequence(CDS)(3±2.30).(4)Bioinformatics analysis:After sequencing and quality control filtering,778643 SNVs and 147859 InDels were obtained.Filtering of rare variants was performed.Only SNVs occurring in exons or splice sites(splicing junction 10 bp)are further analyzed since we are interested in amino acid changes.Finally,15054 SNVs and 1900 InDels were reserved.A total of 428 candidate pathogenic loci and 342 candidate pathogenic genes were obtained by combining the same loci,among which 159 were ACMG pathogenic.Enrichment analysis of candidate genes demonstrated that the KEGG pathways of Porphyrin and chlorophyll metabolism,Ascorbate and aldarate metabolism,Steroid hormone biosynthesis,Pentose and glucuronate interconversions,Retinol metabolism,Drug metabolism,Metabolic pathways showed significant difference between ITP patients and normal controls.Moreover,difference of total exons between ITP patients with abnormal increased megakaryocytes and those with normal was analyzed,as well as the difference of total exons between rhTPO responsive and non-responsive ITP patients.(5)Verification of analysis results:6 SNVs of FHOD1 were randomly selected to carry out Sanger sequencing to confirm our sequencing and analysis results.Sanger sequencing on 200 ITP samples showed the results were rs200317614(0.5%),rs6499118(3.0%),rs117880323(0.5%),rs 199924523(0.5%),rs539519503(0.5%)and 67264047(0.5%).Seven candidate genes were randomly selected to carry out RT-qPCR,and results showed that the expression of MMP9,ERCC6 and NFIB was significantly different between ITP and controls.Conclusions:(1)The genomes of ITP patients were indeed different from those of normal controls.The intron region had the largest number of SNVs,followed by the exon region,and the intron region also had the largest number of InDels,followed by the intergenic region,while the number of CNV variation was very small.(2)A total of 428 candidate pathogenic loci and 342 candidate pathogenic genes were obtained,among which 159 were ACMG pathogenic.Enrichment analysis of candidate genes demonstrated that the KEGG pathways of Porphyrin and chlorophyll metabolism,Ascorbate and aldarate metabolism,Steroid hormone biosynthesis,Pentose and glucuronate interconversions,Retinol metabolism,Drug metabolism,Metabolic pathways showed significant difference between ITP patients and normal controls.(3)The difference of total exons between ITP patients with abnormal increased megakaryocytes and those with normal,as well as the difference of total exons between rhTPO responsive and non-responsive ITP patients,may be of important guiding significance to explore the heterogeneity related to megakaryocytes among ITP patients and to predict the efficacy of rhTPO.Part ?:Investigating the influence of rhTPO on megakaryocyte and platelet production under different antiplatelet antibodies condition in ITPBackground:Primary immune thrombocytopenia(ITP)is an acquired autoimmune disease characterized by a persistent decrease in platelet count.So far,the pathogenesis of ITP has not been fully elucidated and its center hypothesis is "immune intolerance".Humoral and cellular immune abnormalities in the body of ITP accelerated the destruction of platelet.Loss tolerance to platelet-specific autoantibodies and other immune abnormalities also affect the growth,maturation and normal apoptosis of megakaryocyte.The abnormal influence of immune system on the platelet and megakaryocyte contributed to the increased destruction and decreased production of platelet.The specific autoantibodies against glycoprotein(GP)on the surface of platelet membrane are the main molecules that attack platelets,and up to 70%of patients with ITP have platelet-specific autoantibodies,which were generally against GP?b/?a or GP?b/?.The type of epitopes targeted by autoantibodies may influence the therapeutic effect of ITP.As an important part of humoral immunity,antibody is also the earliest discovered pathogenesis of ITP and plays an important part in the pathogenesis of ITP.Several studies have shown that anti-GP?b/?a and anti-GPIb/IX antibodies have different effects on platelet production.In vivo murine passive ITP model,antibodies of different antiplatelet specificity have different biological effects on megakaryocytes.Some anti-GP?b/?a antibodies could inhibit the proliferation of megakaryocytes and cause the number of megakaryocytes decrease,but anti-GPIba antibodies had little effect on the number of megakaryocytes in bone marrow.The knockout of GPIba gene could result in the Bernard-Soulier syndrome in mice which is characterized by severe bleeding and macrothrombocytopenia.GPIba subunit interacts with filamin-1 and 14-3-3zeta,known binding proteins to the GPIba cytoplasmic tail.Interaction between the cytoplasmic domain of GPIb? with its cytoskeletal binding partner,filamin,is a major determinant of platelet size,and deficiency of either protein results in macrothrombocytopenia.In addition to its role in stabilizing GPIb trafficking and function,filamin appears to play a prominent role in regulating platelet size.Filamin A can interact with PACSIN2(A type of BAR/F-BAR protein)to regulate the formation of the inner membrane system of megakaryocytes and the formation of tubular membranous structures in platelets.To sum up,anti-GP?b/?a antibodies mostly block the proliferation of megakaryocytes,and anti-GPIba antibodies mostly affect the internal skeleton structure of platelets and megakaryocytes.Thrombopoietin,produced by the li ver,is the primary regulator of megakaryocyte progenitor expansion and differentiation.RhTPO,like other TPO-RAs,was known to be effective in promoting platelet count and reducing bleeding.However,it has not been clarified whether the dysfunction of megakaryocytopoiesis and thrombocytopoiesis in ITP patients with anti-GP?b/?a antibody positive or anti-GP?b/? antibody positive can be corrected by rhTPO.In the present study,we investigated the proportion,maturation and apoptosis of the megakaryocyte induced from human cord blood CD34+progenitor cells cultured with ITP serum under anti-GP?b/?a antibodies and/or anti-GP?b/? antibodies.To clarify whether the dysfunction of megakaryocytopoiesis and thrombocytopoiesis in ITP patients with anti-GP?b/?a antibody positive or anti-GP?b/? antibody positive can be corrected by rhTPO and to investigate the efficacy of rmTPO on increasing platelet count for murine passive ITP model in vivo.Objectives:(1)To investigate the count,maturation,apoptosis and platelet production of the megakaryocyte induced from human cord blood CD34+progenitor cells cultured with ITP serum under anti-GP?b/?a antibodies and/or anti-GP?b/? antibodies.(2)To clarify whether the dysfunction of megakaryocytopoiesis and thrombocytopoiesis in ITP patients with anti-GP?b/?a antibody positive or anti-GP?b/? antibody positive can be corrected by rhTPO.(3)To investigate the efficacy of rmTPO on increasing platelet count for murine passive ITP model in vivo.Methods:(1)ITP patients and healthy controls:Collecting peripheral blood from 41 ITP patients and 8 controls.Serum and plasma were prepared aseptically.(2)Indirect MAIPA test:The results of two kinds of platelet antibodies in the plasma of ITP patients were detected,and the serum of patients with different MAIPA results were collected.(3)Megakaryocyte culture induced from human umbilical cord blood CD34+progenitor cells:CD34+cells were cultured in serum free expansion medium(SEFM)and were incubated with three kinds of megakaryocyte-induced cytokines(recombinant human thrombopoietin,stem cell factor and recombinant human interleukin 3).(4)In vitro study:Different serum was added in different groups.Part 1:Group A:anti-GP?b/?a antibody positive serum,Group B:anti-GPIb/IX antibody positive serum,Group C:double negative serum,D:healthy serum.Part 2:Group A:double negative serum,Group B:double negative serum+anti-GPIIb antibody,Group C:double negative serum +anti-GPIba antibody,Group D:double negative serum+ anti-GPIIb+anti-GPIb? antibodies.After 14 days culturing,the quantity,polyploidy proportion and apoptosis ratio of megakaryocytes and the number of platelets released were detected.RhTPO was added in 4 groups of part 2.Continuing to culture with rhTPO for 14 days,the count,polyploidy,apoptosis ratio of megakaryocytes and the number of platelets released were detected again.(5)In vivo murine passive ITP model study:Group A:C57BL/6 mice+anti-GPIIb antibody+rmTPO;Group B:C57BL/6 mice+NS;Group C:C57BL/6 mice+anti-GPIb? antibody+rmTPO;Group D:C57BL/6 mice+anti-GPIba antibody+NS.Platelet counts were tested and recorded weekly to evaluate and compare the response of rmTPO on two kinds of murine passive ITP models.Results:(1)There was no significant difference in serum TPO levels between normal controls and ITP patients with different MAIPA results.Forty-one ITP patients and eight normal controls were finally enrolled in our research,and they were divided into five groups according to the results of MAIPA test.Positive anti-platelet antibodies were detected in 28 patients who accounted for 68.3%.There was no significant difference in serum TPO levels between normal controls(n=8),anti-GP?b/?a positive group(n=13),anti-GP?b/? positive group(n=8),double negative group(n=13)and double positive group(n=8)(P>0.05).(2)The effects of serum of normal controls and ITP patients with different MAIPA results on umbilical cord blood induced megakaryocytes and platelet formation were heterogeneous.There was no significant difference in the number of total cells between normal control group(n=8),anti-GP?b/?a positive group(n=13),anti-GP?b/? positive group(n=8)and double negative ITP group(n=13)(P>0.05).The number and proportion of CD61+megakaryocytes were significantly lower in the anti-GP?b/?a positive group and anti-GP?b/? positive group than those in the normal control group.And the anti-GP?b/?a positive group was also significantly lower than the double negative group.The number of platelets released in all ITP groups were significantly lower than that in normal control group.(3)Effects of anti-platelet GP?b? and GP?b antibodies on the quantity,polyploid proportion and apoptosis ratio of megakaryocytes and platelet release were different.The proportion of CD61+megakaryocytes was significantly lower in the anti-GP?b antibody group(n=10)and the double antibodies group(n=10)than that in the antibody-free group(n=10).The proportion of polyploid megakaryocytes in two antibodies group was significantly lower than that in antibody-free group.There was no significant difference in apoptosis ratio between four groups.However,the number of platelets released in all antibody groups were significantly lower than that in antibody-free group.(4)The effects of rhTPO on rescuing impaired megakaryocytopoiesis and thrombocytopoiesis caused by anti-GP?b antibody or anti-GPIba antibody were different.After rhTPO treatment for 14 days,the proportion of CD61+megakaryocytes in antibody-free group and anti-GP?b antibody group was significantly higher than that in the double antibodies group.And the elevated percentage of CD61+megakaryocytes in anti-GP?b antibody group was significantly higher than that in anti-GPIba antibody group(P=0.035).RhTPO could regulate the polyploid proportion of megakaryocytes in double antibodies group to relative normal level.There was still no significant difference in apoptosis ratio between four groups after rhTPO treatment.After rhTPO treatment,the number of platelets released in four groups increased.However,the anti-GPIba antibody group and double antibodies group still had significantly fewer platelet count than the antibody-free group.In contrast,there was no significant difference between the anti-GP?b antibody group and the antibody-free group.(5)Murine passive ITP models caused by anti-GPIba autoantibodies were less responsive to rm TPO treatment than those caused by anti-GP?b antibodies.RmTPO could increase the platelet counts of the murine passive ITP model caused by anti-GP?b antibodies significantly.However,the efficacy of rmTPO on platelet elevating for anti-GP?b? antibodies induced ITP models was poor.Conclusions:(1)There was no significant difference in serum TPO levels between normal controls and ITP patients with different MAIPA results.(2)The effects of serum of normal controls and ITP patients with different MAIPA results on umbilical cord blood-induced megakaryocytes and platelet formation were heterogeneous.(3)The effects of rhTPO on rescuing impaired megakaryocytopoiesis and thrombocytopoiesis caused by anti-GP?b antibody or anti-GPIba antibody might be different.(4)Murine passive ITP models caused by anti-GPIba autoantibodies were less responsive to rmTPO treatment than those caused by anti-GP?b antibodies.(5)MAIPA test may help to predict which patients will benefit more from rhTPO treatment.Part III:Platelet autoantibody specificity and response to rhTPO treatment in patients with primary immune thrombocytopeniaBackground:Primary immune thrombocytopenia(ITP)is an acquired autoimmune bleeding disorder.It occurs in 2?5 per 105 adults each year and leads to variable bleeding symptoms.Corticosteroids and intravenous immunoglobulin(IVIG)remain the major first-line treatment for ITP.However,about 15 to 25%of patients will not respond to first-line treatments.As the pathogenic mechanism of impaired megakaryocytopoiesis and decreased platelet production was revealed in ITP,TPO receptor agonists(TPO-RAs),including romiplostim,eltrombopag and avatrombopag,have been clinically used for the management of ITP,with the efficacy of 60%to 90%.rhTPO,a full-length glycosylated TPO,showed similar efficacy as other TPO-RAs in corticosteroid-resistant or relapsed ITP patients.However,rhTPO remains a high-cost medication for ITP patients.Anti-glycoprotein(GP)autoantibodies,which were generally against GP?b/?a or GPIb/IX,could be serve as predictive indicators for several treatments in ITP.It has been reported that ITP patients with anti-GPIb/IX were less responsive to IVIG and corticosteroids,while patients with anti-GP?b/?a autoantibodies were more responsive to rituximab.In the present study,we retrospectively analyzed the relationship between rhTPO efficacy and plasma autoantibody specificities in adult ITP patients relapsed or resistant to first-line treatments,and found that anti-GPIb/IX positive patients were less responsive to rhTPO.Objective:Retrospective analysis was performed on ITP patients who had received rhTPO,failed to respond to first-line treatment or relapsed,and had detected plasma platelet autoantibodies.Statistical analysis was performed to explore the correlation between the efficacy of rhTPO and the specificity of autoantibodies,aiming to find indicators that could predict the efficacy of rhTPO,so as to be beneficial to the individualized and precise treatment of ITP.Methods:(1)Study patients:We retrospectively reviewed the clinical records of hospitalized adult ITP patients who underwent rhTPO therapy and had platelet autoantibody tests performed before rhTPO treatment between August 2010 and April 2019 at the Department of Hematology,Qilu Hospital,Shandong University.All included patients met the International Working Group diagnostic criteria for ITP and had a history of no response or relapsed after first-line treatments.Previous therapies must be completed at least 4 weeks before rhTPO administration.Patients complicated by pregnancy were excluded.Approval for the study was obtained from the ethics committee of Qilu Hospital,Shandong University.(2)Study variables:Plasma anti-GP?b/?a and anti-GP?b/? autoantibodies were determined by modified monoclonal antibody specific immobilization of platelet antigen(MAIPA)assay before treatment.rhTPO was injected subcutaneously at a dose of 300 U/kg once daily for 14 days.When platelet counts remained less than 30 × 109/L or bleeding symptoms were still present after 14 days of rhTPO injection,treatment was considered unsuccessful.Time to response(TTR),baseline platelet count,disease duration,MAIPA results and incidence of response were reviewed.Overall response(OR)was defined as platelet counts rose above 30 × 109/L and at least twice the baseline amount on 2 separate occasions which were at least 7 days apart without bleeding symptoms according to the proposals of the International Working Group.Complete response(CR)was defined as platelet counts rose above 100 × 109/L by the same criteria.(3)Statistical analysis:Statistical analyses were conducted using Student t test,one-way ANOVA,Student-Newman-Keuls test,or Mann-Whitney U test for quantitative variables and ?2 tests for categorical variables.Bonferroni's correction was used for multigroup comparisons.Multivariable logistic regression and factorial design were used to estimate the influence factors on treatment response,and results were expressed as odds ratio and 95%confidence intervals(CIs).All P values were two-sided,and statistical significance was defined as a P value of<0.05.All statistical analyses were performed with SPSS 23.0.Results:(1)According to inclusion and exclusion criteria,123 ITP patients were eligible for analysis,Among the patients included,47.2%(58/123)had autoantibodies against GP?b/?a,40.6%(50/123)had autoantibodies against GP?b/?,30.9%(38/123)had autoantibodies against both GP?b/?a and GP?b/?(double positive),and 43.1%(53/123)had neither anti-GP?b/?a nor anti-GP?b/? autoantibodies(double negative).(2)73.2%(90/123)of patients responded to rhTPO treatment,including 53 with CR and 37 with partial response.For the 90 patients with initial response,the median TTR was 7 days(range:2-13 days).60.0%(30/50)of patients with anti-GP?b/?autoantibodies achieved response,as compared with 82.2%(60/73)of patients who were negative for anti-GP?b/? autoantibodies(?2=7.444,P=0.006),Statistical difference in CR rate was also found between patients with and without anti-GP?b/?autoantibodies(36.0%vs.47.9%,x2=5.448,P=0.020).We did not observe any statistical difference in OR or CR rate between patients with anti-GP?b/?a autoantibodies and those without(all P>0.05).The OR rate showed significant difference among anti-GP?b/? single-positive group(33.3%),double positive group(68.4%),double negative group(81.1%)and anti-GP?b/?a single-positive group(85.0%;P=0.004).Further comparison showed that anti-GP?b/? single-positive patients was more resistant to rhTPO treatment than anti-GP?b/?a single-positive cases(P=0.006).Nevertheless,statistical significance was not achieved in OR rate between double-positive patients and double-negative patients or anti-GP?b/?a single-positive patients after Bonferroni correction(all P>0.05/6).Additionally,there was no statistical difference in CR rate among 4 groups(P=0.065).(3)Multivariate logistic regression analysis showed that age,gender,baseline platelet count or anti-GP?b/?a autoantibody were not associated with patients' response to rhTPO treatment.Only the presence of anti-GP?b/? autoantibodies was related to patients' response rate in the logistic regression model with stepwise method.The odd ratio of response for patients with anti-GP?b/? autoantibodies versus those without anti-GP?b/? autoantibodies was 0.040(P=0.003,95%CI:0.005-0.336).The 2 × 2 factorial design analysis did not find interaction effect between autoantibody species on response either,and anti-GP?b/? autoantibodies remained the only factor which had obvious effect on the OR rate(F=12.826,P<0.001).Therefore,the presence of anti-GP?b/? autoantibodies might be a predictive marker for poor response to rhTPO treatment.Conclusions:(1)Plasma anti-GP?b/? positive patients tested by MAIPA showed poor response to rhTPO treatment,while anti-GP?b/?a antibodies showed no predictive effect.(2)There was no interaction effect between anti-GP?b/? antibodies and anti-GP?b/?a antibodies on the efficacy of rhTPO for ITP.(3)This study suggested that rhTPO could be recommended in patients with negative anti-GP?b/? antibodies,and this group of patients may benefit more.