Dipartimento Di Ingegneria Industriale E Dell'Informazione, Università Degli Studi Di Pavia, 27100, Pavia, Italy.
Consultant, Milan, Italy.
AAPS J. 2024 Aug 8;26(5):92. doi: 10.1208/s12248-024-00960-4.
Tumor volume doubling time (TVDT) has been shown to be a potential surrogate marker of biological tumor activity. However, its availability in clinics is strongly limited due to ethical and practical reasons, as its assessment requires at least two subsequent tumor volume measurements in untreated patients. Here, a translational modeling framework to predict TVDT distributions in untreated cancer patient populations from tumor growth data in patient-derived xenograft (PDX) mice is proposed. Eleven solid cancer types were considered. For each of them, a set of tumor growth studies in PDX mice was selected and analyzed through a mathematical model to characterize the distribution of the exponential tumor growth rate in mice. Then, assuming an exponential growth of the tumor mass in humans, the growth rates were scaled from PDX mice to humans through an allometric scaling approach and used to predict TVDTs in untreated patients. A very good agreement was found between model predicted and clinically observed TVDTs, with 91% of the predicted TVDT medians fell within 1.5-fold of observations. Further, exploiting the intrinsic relationship between tumor growth dynamics and progression free survival (PFS), the exponential growth rates in humans were used to generate the expected PFS curves in absence of anticancer treatment. Predicted curves were extremely close to published PFS data from studies involving patient cohorts treated with supportive care or low effective therapies. The proposed approach shows promise as a potential tool to increase knowledge about TVDT in humans without the need of directly measuring tumor dimensions in untreated patients, and to predict PFS curves in untreated patients, that could fill the absence of placebo-controlled arms against which to compare treaded arms during clinical trials. However, further validation and refinement are needed to fully assess its effectiveness in this regard.
肿瘤体积倍增时间（TVDT）已被证明是生物肿瘤活性的潜在替代标志物。然而，由于伦理和实际原因，其在临床上的应用受到很大限制，因为其评估需要在未经治疗的患者中至少进行两次后续的肿瘤体积测量。在这里，提出了一种从患者来源的异种移植（PDX）小鼠的肿瘤生长数据预测未经治疗的癌症患者人群中 TVDT 分布的转化建模框架。考虑了 11 种实体癌类型。对于每一种类型，选择了一组 PDX 小鼠中的肿瘤生长研究，并通过数学模型进行分析，以表征小鼠中肿瘤生长率的分布。然后，假设肿瘤质量在人体内呈指数增长，通过一种异速缩放方法将生长率从 PDX 小鼠缩放至人类，并用于预测未经治疗患者的 TVDT。模型预测的 TVDT 中位数与临床观察的 TVDT 中位数非常吻合，91%的预测 TVDT 中位数在观察值的 1.5 倍以内。此外，利用肿瘤生长动力学与无进展生存期（PFS）之间的内在关系，人类的指数增长率被用于在没有抗癌治疗的情况下生成预期的 PFS 曲线。预测曲线与涉及接受支持性治疗或低有效治疗的患者队列的研究中发表的 PFS 数据非常接近。所提出的方法有望成为一种潜在的工具，无需直接测量未经治疗患者的肿瘤尺寸，即可增加对人类 TVDT 的了解，并预测未经治疗患者的 PFS 曲线，这可以填补临床试验中缺乏安慰剂对照臂来比较治疗臂的空白。然而，需要进一步的验证和改进来充分评估其在这方面的有效性。
