Medical advances including the neoadjuvant anti-PD-1 immunotherapy play a role in promoting clinical outcomes such as improved overall and progression-free survival probabilities. This paper considers the regression analysis of current status data with a cured subgroup in the population using a semiparametric non-mixture cure model. We propose a sieve maximum likelihood estimation for the model with the Bernstein polynomials. Moreover, an expectation–maximization (EM) algorithm is developed under the non-mixture cure model to calculate the estimators for both parametric and non-parametric components. Under some mild conditions, the asymptotic properties of the estimators are established, including the strong consistency, the convergence rate and the asymptotic normality. Simulation studies are conducted to investigate the finite sample performance of the proposed estimators. A real dataset from the tumorigenicity experiment is analysed for illustration.

中文翻译：

---

使用当前状态数据对非混合物固化模型进行有效估计

包括新辅助抗 PD-1 免疫疗法在内的医学进步在促进临床结果方面发挥作用，例如提高总体和无进展生存概率。本文考虑使用半参数非混合物治愈模型对人群中治愈亚组的现状数据进行回归分析。我们为具有伯恩斯坦多项式的模型提出了筛子最大似然估计。此外，在非混合物固化模型下开发了一种期望最大化 (EM) 算法来计算参数和非参数组件的估计量。在一些温和的条件下，建立了估计量的渐近性质，包括强一致性、收敛速度和渐近正态性。进行模拟研究以调查所提出的估计器的有限样本性能。分析来自致瘤性实验的真实数据集以进行说明。