Digital Biomarkers for Oncology Group, National Center for Tumor Diseases (NCT), German Cancer Research Center (DKFZ), Heidelberg, Germany.
Department of Urology and Urological Surgery, Medical Faculty Mannheim of Heidelberg University, University Medical Center Mannheim, Mannheim, Germany.
BJU Int. 2021 Sep;128(3):352-360. doi: 10.1111/bju.15386. Epub 2021 May 5.
To develop a new digital biomarker based on the analysis of primary tumour tissue by a convolutional neural network (CNN) to predict lymph node metastasis (LNM) in a cohort matched for already established risk factors.
Haematoxylin and eosin (H&E) stained primary tumour slides from 218 patients (102 N+; 116 N0), matched for Gleason score, tumour size, venous invasion, perineural invasion and age, who underwent radical prostatectomy were selected to train a CNN and evaluate its ability to predict LN status.
With 10 models trained with the same data, a mean area under the receiver operating characteristic curve (AUROC) of 0.68 (95% confidence interval [CI] 0.678-0.682) and a mean balanced accuracy of 61.37% (95% CI 60.05-62.69%) was achieved. The mean sensitivity and specificity was 53.09% (95% CI 49.77-56.41%) and 69.65% (95% CI 68.21-71.1%), respectively. These results were confirmed via cross-validation. The probability score for LNM prediction was significantly higher on image sections from N+ samples (mean [SD] N+ probability score 0.58 [0.17] vs 0.47 [0.15] N0 probability score, P = 0.002). In multivariable analysis, the probability score of the CNN (odds ratio [OR] 1.04 per percentage probability, 95% CI 1.02-1.08; P = 0.04) and lymphovascular invasion (OR 11.73, 95% CI 3.96-35.7; P < 0.001) proved to be independent predictors for LNM.
In our present study, CNN-based image analyses showed promising results as a potential novel low-cost method to extract relevant prognostic information directly from H&E histology to predict the LN status of patients with prostate cancer. Our ubiquitously available technique might contribute to an improved LN status prediction.
开发一种基于卷积神经网络（CNN）分析原发肿瘤组织的新数字生物标志物，以预测已建立的危险因素匹配队列中淋巴结转移（LNM）的情况。
选择 218 例接受根治性前列腺切除术的患者（102 例 N+；116 例 N0）的苏木精和伊红（H&E）染色原发肿瘤切片，这些患者的 Gleason 评分、肿瘤大小、静脉侵犯、神经周围侵犯和年龄匹配，以训练 CNN 并评估其预测 LN 状态的能力。
使用相同数据训练 10 个模型，获得了平均受试者工作特征曲线下面积（AUROC）为 0.68（95%置信区间 [CI] 0.678-0.682）和平均平衡准确率为 61.37%（95%CI 60.05-62.69%）。平均敏感度和特异性分别为 53.09%（95%CI 49.77-56.41%）和 69.65%（95%CI 68.21-71.1%）。这些结果通过交叉验证得到了证实。在 N+样本的图像切片上，LNM 预测的概率评分明显更高（平均[N+]概率评分 0.58 [0.17]与[N0]概率评分 0.47 [0.15]，P = 0.002）。在多变量分析中，CNN 的概率评分（每百分比概率的优势比[OR]为 1.04，95%CI 1.02-1.08；P = 0.04）和血管淋巴管侵犯（OR 11.73，95%CI 3.96-35.7；P < 0.001）被证明是 LNM 的独立预测因子。
在本研究中，基于 CNN 的图像分析显示出有希望的结果，作为一种从 H&E 组织学中直接提取相关预后信息以预测前列腺癌患者 LN 状态的潜在低成本方法。我们普遍可用的技术可能有助于改善 LN 状态预测。
