磁共振波谱检测三阴性乳腺癌患者尿液代谢生物标志物预测化疗敏感性的临床研究

Screening of urinary metabolic biomarkers for predicting chemotherapy sensitivity in patients with triple-negative breast can-cer by magnetic resonance spectroscopy

目的通过磁共振波谱(NMR)代谢组学方法探讨不同化疗敏感性的三阴性乳腺癌(TNBC)患者尿液中的差异性代谢生物标志物,为早期预测患者新辅助化疗(NAC)敏感性提供依据。方法选取2019年5月至2020年12月在浙江省肿瘤医院接受NAC的TNBC患者52例,根据疗效分为病理完全缓解(pCR)患者8例(pCR组),病理部分缓解(pPR)患者28例(p PR组),病理稳定(pSD)患者16例(pSD组)。通过NMR测定3组患者尿液的代谢物谱图并进行多元统计分析。根据K-近邻算法(KNN)分类判别,寻找潜在的判别标志物,并利用ROC曲线计算AUC进行评估。结果3组患者年龄、BMI、Ki-67、月经状态和肿瘤分期比较差异均无统计学意义(均P>0.05)。根据患者尿液NMR代谢谱图并结合主成分分析,发现3组患者尿液的代谢表型有所不同。根据KNN分类判别,在尿液寻找出19种代谢物作为潜在的生物标志物。根据ROC的AUC值评估,发现蛋氨酸是潜在的生物标志物,预测pCR组、pPR组和pSD组TNBC化疗敏感性的AUC分别为0.8324、0.9539和0.8264(均P<0.05)。结论不同化疗敏感性的TNBC患者具有不同的尿液代谢表型,其中蛋氨酸可以作为区别不同化疗敏感性患者的代谢生物标志物。

Objective To screening metabolic biomarkers in the urine of patients with triple-negative breast cancer(TNBC)for predicting chemotherapy sensitivity based on nuclear magnetic resonance(NMR).Methods Fifty-two patients with TNBC undergoing neoadjuvant chemotherapy from May 2019 to December 2020 at Zhejiang Cancer Hospital were enrolled.There were 8 cases with complete pathological remission(pCR group),28 cases with pathological partial remission(pPR group)and 16 cases with pathological stable disease(pSD group)after chemotherapy.The urinary metabolite profiles were measured by NMR and subjected to multivariate statistical analysis.Potential discriminatory markers were screened according to the K-nearest neighbor(KNN)classification and assessed by ROC curves.Results There were no significant differences in age,BMI,Ki 67,menstrual status and tumor stage among three groups of patients.Based on the urinary metabolic profiles and principal component analysis,the differentially expressed metabolic phenotypes among the three groups were screened.Nineteen metabolites were identified as potential urinary biomarkers according to the KNN classification.The area under ROC curve(AUC)of methionine as a potential biomarker of chemosensitivity in pCR,pPR and pSD groups was 0.8324,0.9539 and 0.8264,respectively.Conclusion The study shows that the urinary methionine level may be used as a potential metabolic biomarker for chemotherapy sensitivities in TNBC patients undergoing neoadjuvant chemotherapy.