基于MRI和微小RNA-7106-3p模型评价进展期直肠癌新辅助放化疗的疗效:127例前瞻性临床试验

  目的  构建基于MRI和微小核糖核酸-7106-3p（miR-7106-3p）的列线图回归模型，探究其评价进展期直肠癌新辅助放化疗（NCRT）疗效的价值。  方法  前瞻性纳入2019年3月~2022年2月于河北省退役军人总医院就诊的直肠癌患者127例，NCRT后行全直肠系膜切除术。NCRT前1周内行MRI检查并抽取患者外周血，采用实时荧光定量聚合酶链反应法检测血清miR-7106-3p水平。进展期直肠癌NCRT的风险因素用Logistic回归分析。构建风险因素的列线图回归模型，采用决策曲线分析法、一致性指数和校准曲线评价模型进行分析。  结果  37例患者（29.13%）经NCRT后达到病理学完全缓解（pCR）（pCR组），90例（70.87%）未达到pCR（非pCR组）。pCR组的容积转运常数（Ktrans）、血管外细胞外间隙容积比（V_e）、回流速率常数（K_ep）和miR-7106-3p水平均高于非pCR组，表观弥散系数（ADC）低于非pCR组（P < 0.05）。Logistic回归分析结果显示，N分期越高、ADC > 0.92×10-3 mm2/s是进展期直肠癌NCRT后pCR的独立危险因素（P < 0.05），Ktrans > 0.33 min、V_e > 0.55%、K_ep > 0.54 min和miR-7106-3p > 0.31是进展期直肠癌NCRT后pCR的独立保护因素（P < 0.05）。模型C（由N分期、Ktrans、V_e、K_ep、ADC和miR-7106-3p构成）的一致性指数为0.977，高于模型A（由Ktrans、V_e、K_ep和ADC构成，0.957）和模型B（由N分期、Ktrans、V_e、K_ep和ADC构成，0.956）。模型B的平均绝对误差为0.015，低于模型A（0.017）和模型C（0.024）。阈值概率在0.10~1.0的大部分范围内，模型C的净收益高于模型A和模型B。  结论  基于MRI和miR-7106-3p的模型能较好的评价进展期直肠癌NCRT疗效。 

Value of the model based on MRI and microRNA-7106-3p for evaluating the efficacy of neoadjuvant chemoradiotherapy in advanced rectal cancer: a prospective study based on 127 patients

  Objective  To construct a nomogram regression model based on MRI and microRNA-7106-3p (miR-7106-3p), and investigate its value in evaluating the efficacy of neoadjuvant chemoradiotherapy (NCRT) for advanced rectal cancer.  Methods  A total of 127 patients with rectal cancer were prospectively enrolled from March 2019 to February 2022 at Hebei General Hospital For Veterans who underwent total rectal mesenteric resection after NCRT. MRI was performed within 1 week before NCRT. Peripheral blood was extracted from patients at the same time. The serum miR-7106-3p level was detected by real-time fluorescence quantitative polymerase chain reaction. Risk factors for the efficacy of NCRT in advanced rectal cancer were analysed by Logistic regression. The nomogram regression model for risk factors was constructed. Consistency index, calibration curve and decision curve analysis were used to evaluate the value of the model.  Results  There were 37 (29.13%) patients achieved pathological complete remission (pCR) after NCRT (pCR group), and 90 patients (70.87%) did not achieved pCR (non-pCR group). The levels of volume transport constant (Ktrans), extravascular extracellular space volume ratio (V_e), reflux rate constant (K_ep) and miR-7106-3p were higher in the pCR group than in the non-pCR group, and apparent dispersion coefficient (ADC) was lower than in the non-pCR group (P < 0.05). The logistic regression analysis showed that higher N stage and ADC > 0.92×10-3 mm2/s were independent risk factors for pCR after NCRT in progressive rectal cancer (P < 0.05), and Ktrans > 0.33 min, V_e > 0.55%, K_ep > 0.54 min and miR-7106-3p > 0.31 were independent protective factors for pCR after NCRT in progressive rectal cancer (P < 0.05). The consistency index of model C (consisting of N-stage, Ktrans, Ve, Kep, ADC and miR-7106-3p) was 0.977, which was higher than that of model A (consisting of Ktrans, V_e, K_ep and ADC, 0.957) and model B (consisting of N-stage, Ktrans, V_e, K_ep and ADC, 0.956). The mean absolute error of model B was 0.015, which was lower than that of model A (0.017) and model C (0.024). When the threshold probabilities ranged from 0.10-1.0 for most of the range, the net benefit of model C was higher than that of model A and model B.  Conclusion  The model can better evaluate the efficacy of NCRT in progressive rectal cancer based on MRI and miR-7106-3p.