Deep learning model based on primary tumor to predict lymph node status in clinical stage IA lung adenocarcinoma:a multicenter study

Objective:To develop a deep learning model to predict lymph node(LN)status in clinical stage IA lung adeno-carcinoma patients.Methods:This diagnostic study included 1,009 patients with pathologically confirmed clinical stage T1N0M0 lung adenocarcinoma from two independent datasets(699 from Cancer Hospital of Chinese Academy of Medical Sciences and 310 from PLA General Hospital)between January 2005 and December 2019.The Cancer Hospital dataset was randomly split into a training cohort(559 patients)and a validation cohort(140 patients)to train and tune a deep learning model based on a deep residual network(ResNet).The PLA Hospital dataset was used as a testing cohort to evaluate the generalization ability of the model.Thoracic radiologists manually segmented tumors and interpreted high-resolution computed tomography(HRCT)features for the model.The predictive performance was assessed by area under the curves(AUCs),accuracy,precision,recall,and F1 score.Subgroup analysis was performed to evaluate the potential bias of the study population.Results:A total of 1,009 patients were included in this study;409(40.5%)were male and 600(59.5%)were female.The median age was 57.0 years(inter-quartile range,IQR:50.0-64.0).The deep learning model achieved AUCs of 0.906(95%CI:0.873-0.938)and 0.893(95%CI:0.857-0.930)for predicting pN0 disease in the testing cohort and a non-pure ground glass nodule(non-pGGN)testing cohort,respectively.No significant difference was detected between the testing cohort and the non-pGGN testing cohort(P=0.622).The precisions of this model for predicting pN0 disease were 0.979(95%CI:0.963-0.995)and 0.983(95%CI:0.967-0.998)in the testing cohort and the non-pGGN testing cohort,respectively.The deep learning model achieved AUCs of 0.848(95%CI:0.798-0.898)and 0.831(95%CI:0.776-0.887)for predicting pN2 disease in the testing cohort and the non-pGGN testing cohort,respectively.No significant difference was detected between the testing cohort and the non-pGGN testing cohort(P=0.657).The recalls of this model for predicting pN2 disease were 0.903(95%CI:0.870-0.936)and 0.931(95%CI:0.901-0.961)in the testing cohort and the non-pGGN testing cohort,respectively.Conclusions:The superior performance of the deep learning model will help to target the extension of lymph node dissection and reduce the ineffective lymph node dissection in early-stage lung adenocarcinoma patients.

Number of involved nodal stations: a better lymph node classification for clinical stage IA lung adenocarcinoma

Background:With the popularization of lung cancer screening,more early-stage lung cancers are being detected.This study aims to compare three types of N classifications,including location-based N classification(pathologic nodal classification[pN]),the number of lymph node stations(nS)-based N classification(nS classification),and the combined approach proposed by the International Association for the Study of Lung Cancer(IASLC)which incorporates both pN and nS classification to determine if the nS classification is more appropriate for early-stage lung cancer.Methods:We retrospectively reviewed the clinical data of lung cancer patients treated at the Cancer Hospital,Chinese Academy of Medical Sciences between 2005 and 2018.Inclusion criteria was clinical stage IA lung adenocarcinoma patients who underwent resection during this period.Sub-analyses were performed for the three types of N classifications.The optimal cutoffvalues for nS classification were determined with X-tile software.Kaplan‒Meier and multivariate Cox analyses were performed to assess the prognostic significance of the different N classifications.The prediction performance among the three types of N classifications was compared using the concordance index(C-index)and decision curve analysis(DCA).Results:Of the 669 patients evaluated,534 had pathological stage N0 disease(79.8%),82 had N1 disease(12.3%)and 53 had N2 disease(7.9%).Multivariate Cox analysis indicated that all three types of N classifications were independent prognostic factors for prognosis(all P<0.001).However,the prognosis overlaps between pN(N1 and N2,P=0.052)and IASLC-proposed N classification(N1b and N2a1[P=0.407],N2a1 and N2a2[P=0.364],and N2a2 and N2b[P=0.779]),except for nS classification subgroups(nS0 and nS1[P<0.001]and nS1 and nS>1[P=0.006]).There was no significant difference in the C-index values between the three N classifications(P=0.370).The DCA results demonstrated that the nS classification provided greater clinical utility.Conclusion:The nS classification might be a better choice for nodal classification in clinical stage IA lung adeno-carcinoma.

利用抗原结合多肽嫁接抗体技术制备抗hCG单域抗体

本研究旨在人绒毛膜促性腺激素(hCG)的结合多肽的基础上应用嫁接抗体技术制备抗hCG单域抗体,简化单域抗体制备过程,提高多肽生化稳定性。利用单域抗体通用骨架(cAbBCII10),以hCG结合多肽取代互补决定区CDR1或CDR3,合成cAb BCII10嫁接抗体全基因序列并与sfGFP基因序列融合后,插入到带有His标签的原核表达载体pET30a(+)中,成功构建了pET30a-(His6)-cAbBCII10-CDR1/hCGBP1-sfGFP与pET30a-(His6)-cAbBCII10-CDR3/hCGBP3-sfGFP融合蛋白表达质粒。将重组质粒转化大肠杆菌BL21(DE3),用IPTG诱导表达融合蛋白,得到高表达量的可溶性融合蛋白。利用Ni-NTA亲和柱纯化得到纯蛋白,应用SDS-PAGE鉴定纯化的蛋白为正确表达的目标蛋白。通过抗原抗体结合实验,发现hCG结合多肽嫁接到单域抗体通用骨架的互补决定区CDR1或CDR3后都有抗原结合活性,具有相似的抗体滴度,且嫁接到CDR3后的抗原结合活性比CDR1要高(2–3倍)。嫁接抗体基本保留了所用单域抗体框架较为稳定的生化特性,具有一定的热稳定性和较好的碱耐受性,同时,所接入的hCG结合片段对hCG具有较特异的结合活性。

应用RAD多肽展示体系制备抗hCG类抗体分子

应用基于激烈火球菌Pyrococcus furiosus重组酶RadA的ATP酶结构域(RAD骨架)的多肽展示体系,通过嫁接人绒毛膜促性腺激素(hCG)结合多肽,制备抗hCG类抗体分子。通过合成hCG结合多肽插入RAD多肽展示位点的类抗体基因,成功构建了pET30a-RAD/hCGBP-sfGFP原核表达载体,在大肠杆菌中诱导蛋白表达,分离、纯化获得类抗体蛋白,通过亲和吸附-GFP荧光检测方法测定类抗体对hCG的结合活性,并与应用单域抗体通用骨架制备的嫁接抗体比较活性差异。结果显示,RAD类抗体分子对hCG分子具有较高的亲和性和特异性,显著优于单域嫁接抗体,并与商业单克隆抗体的活性相当;同时,利用RAD多肽展示骨架制备的抗hCG类抗体,具有较高的生化稳定性,是一种具有应用潜力的抗体替代分子。