Yong Zang

A family of Bayesian generalized phase 1–2 oncology trial designs recently was proposed that uses a phase 1–2 design based on early response and toxicity to choose a set of candidate doses rather than one dose, randomizes additional patients among the candidates, and selects a best dose based on progression-free survival time over longer follow-up. This was extended to a generalized phase 1–2–3 design by including possible expansion to a phase 3 trial of the new agent, at its selected optimal dose, compared to standard of care. The decision of whether to conduct phase 3 was based on the predictive probability of a positive trial. This approach was shown to be greatly superior to the conventional three-phase paradigm. In this paper we extend the generalized phase 1–2–3 design to accommodate patient subgroups by doing subgroup-specific dose-finding and phase 3 treatment comparisons. The proposed design adaptively clusters subgroups having similar estimated dose-outcome distributions. While applying this proposed design requires substantial planning, trial conduct is very similar to running a phase 1–2 trial followed by a phase 3 trial. A simulation study shows that the proposed design has much more desirable properties than either ignoring subgroups or running separate trials within subgroups.

Clinical syndromes or diseases with complex etiology often require multiple treatments, each addressing one specific aspect of the disease. Such treatments are typically administered sequentially in an integrated fashion. For example, in alcohol‐associated hepatitis (AH), an acute liver disease caused by excessive alcohol drinking, successful treatment requires therapies that reduce liver inflammation during the acute phase, followed by interventions that address the underlying alcohol use disorder (AUD) to achieve optimal outcomes. Selecting an optimal treatment combination, however, presents great challenges in study design due to the involvement of multiple drug combinations and treatment phases. In this paper, we describe a new trial design based on a Bayesian model that connects the outcomes from the acute and subsequent treatment phases. We refer to this design as the Bayesian Optimal Adaptive Design for Sequentially Integrated Therapies (BIT) design. BIT incorporates multiple interim analyses with adaptive stopping rules for both futility and superiority, enhancing efficiency while controlling the family‐wise type I error rate and maximizing statistical power. We illustrate the use of BIT design in a clinical trial comparing the efficacy of integrated therapy for severe AH, with sample size determination and design parameter optimization. Simulation studies show that the design possesses excellent operating characteristics, including experiment‐wise type I error rate control. Although the design's development is motivated by studies of AH treatment, the design framework is broadly applicable to other complex diseases requiring sequentially delivered therapies. R codes are provided for BIT implementation.

Supplementary Table S1. The outcomes of all 46 clinical trials with racial comparisons. Supplementary Table S2. The list of clinically actionable genes. Supplementary Table S3. Summary of TCGA cancer types, patient samples and data types surveyed in this study. For each data type and each cancer type, the numbers of EUR, EAS and AFR patients available in the analysis are shown. NA: the data type in that cancer type was not included in the analysis. Cancer types are marked red for further analysis. Supplementary Table S4. Detailed information of independent mutation datasets Supplementary Table S5. Summary of TCGA cancer types with imputed drug data in this study. For each data type and each cancer type, the numbers of EUR, EAS and AFR patients available in the analysis are shown. NA: the data type in that cancer type was not included in the analysis. Cancer types are marked red for further analysis. Supplementary Table S6. FDA approved anti-cancer drugs during 1949 to 2023. Supplementary Table S7. Detailed information of independent cohorts.

Supplementary Fig. S1. The underrepresentation of Hispanic or Latino populations in clinical trials. (A) The summary of all clinical trials in Clinical Trials obtained on March 24, 2024. (B-C) The populations of Hispanic or Latino, non-Hispanic or Latino individuals in clinical trials in different phases (B) and (C) different years. Supplementary Fig. S2. The multi-omics molecular variations of clinically actionable genes. (A) The number of multi-omics molecular alterations of clinically actionable genes across ancestries. (B) The difference of EGFR mutation frequency in White vs. Asian patients with NSCLC. (C) The sample sizes of different racial populations across 23 metastatic cancer types with sample sizes ≥10 for at least two of White, Black, and Asian in PMID: 35120664 cohort. The y-axis denotes the sample size; the x-axis denotes the cancer types; the colors represent the different racial populations. (D) The mutation frequency differences of EGFR in White vs. Asian patients with mNSCLC. (E) The difference of PIK3CA mutation frequency in EUR vs. AFR patients with CRC. One-sided Fisher's exact tests were used for the estimation of the significance of the differences in proportions between two groups. Abbreviations: BLCA: Bladder Squamous cell carcinoma, BRCA: Breast invasive carcinoma, CESC: Cervical squamous cell carcinoma and endocervical adenocarcinoma, COAD: Colon adenocarcinoma, CRC: Colorectal cancer, ESCA: Esophageal carcinoma, HNSC: Head and neck squamous cell carcinoma, KIRC: Kidney renal clear cell carcinoma; LIHC: Liver hepatocellular carcinoma; LUAD: Lung adenocarcinoma; LUSC: Lung squamous cell carcinoma, MESO: Mesothelioma, NSCLC: Non-small cell lung cancer, OV: Ovarian serous cystadenocarcinoma, PAAD: Pancreatic adenocarcinoma, PRAD: Prostate adenocarcinoma, READ: Rectum adenocarcinoma, SARC: Sarcoma, SKCM: Skin cutaneous melanoma, STAD: Stomach adenocarcinoma, THCA: Thyroid carcinoma, UCEC: Uterine corpus endometrial carcinoma, UCS: Uterine carcinosarcoma. Supplementary Fig. S3. Landscape of drug response across multiple cancer types. (A) The landscape of ancestry-associated drug response of 359 drugs or small molecules across multiple cancer types. (B) The difference in imputed drug response to the afatinib in EUR vs. EAS patients with NSCLC. The y-axis denotes imputed ln(IC50). The boxes show the median ±1 quartile, with whiskers extending from the hinge to the smallest or largest value within 1.5 interquartile range from the box boundaries. FDR indicates the adjusted two-sided P value calculated from the propensity score algorithm. (C) The mutation frequency differences of EGFR in White vs. Asian patients with NSCLC. (D) The mutation frequency differences of ERBB4 in White vs. Asian patients with NSCLC. (E) The difference in imputed drug response to the lapatinib in EUR vs. EAS patients with BRCA. (F) The mutation frequency differences of ERBB2 in White vs. Asian patients with BRCA. (G) The difference in imputed drug response to the paclitaxel in EUR vs. EAS patients with STAD. (H) The gene set enrichment analysis of microtubule in ERU and EAS patients with STAD. FDR <0.25 is significant. Supplementary Fig. S4. The difference of ancestry-associated immune characteristics across multiple cancer types. (A) The gene set enrichment analysis of 16 immune pathways across different ancestral populations. (B) The differences of 6 stimulatory immune T cells in different ancestries. (C) The multi-omics molecular differences of immune checkpoint genes across ancestry in multiple cancer types within at least 7 comparisons. Supplementary Fig. S5. The differences of immune features and ICB response in BRCA and NSCLC. (A & B) The difference in lymphocyte infiltration scores (A), and indel neoantigens (B) between EUR and AFR patients with PRAD. (C-H) The difference in TCR richness (C), TCR Shannon (D), lymphocyte infiltration score (E), indel neoantigens (F), SNV-derived neoantigens (G) and TGF beta response (H) between EUR and EAS patients with NSCLC. (I) The difference of TMB between EUR and EAS patients with mNSCLC. Student's t-test was used for the estimation of the significance of the difference in proportions between two groups. Supplementary Fig. S6. The differences of immune features and ICB response in BLCA. (A-F) The difference of CYT (A), GEP (B), TMB (C), lymphocyte infiltration score (D), aneuploidy score (E), and SNV neoantigens (F) in EUR and EAS patients with BLCA. (G-I) The differences of immune checkpoint CTLA4 (G), PDCD1 (H) and CD274 (encoding PD-L1; I) in EUR vs. EAS with BLCA. (J) The difference of response to anti-PD-L1 avelumab in White vs. Asian in BLCA. One-sided Fisher's exact tests were used to estimate the difference significance of the difference in proportions between two groups. Supplementary Fig. S7. The difference of multi-omics molecular variations related to PROTACs. (A & B & C) The number of multi-omics molecular alterations of CART targets (A), PROTAC targets (B) and E3 ligase genes (C) across ancestries. (D) The multi-omics molecular alterations of E3 ligase genes within at least 10 comparisons.

Systematic multiomic analysis has revealed ancestry-dependent molecular alterations, but their impact on the efficacy of anticancer treatment is yet largely unknown. In this study, we analyzed clinical trials from ClinicalTrials.gov and found that only 8,779/102,721 (8.5%) oncology clinical trials posted information on enrollment by race/ethnicity. The underrepresentation of non-White populations suggests that it remains challenging to determine differences in the efficacy of antitumor treatments among different racial groups. Through a comprehensive analysis of clinically actionable genes, imputed drug responses, and immune features, we identified potential differences in treatment responses to targeted therapy, chemotherapy, and immunotherapy between different ancestral populations. Further analysis of multiple independent cohorts confirmed some of our key findings. Such potential ancestral effects are also identified in response to emerging new treatments like chimeric antigen receptor T-cell therapy and proteolysis-targeting chimeras. These findings are made publicly available in a comprehensive web portal, Ancestral Differences of Efficacy in Cancers (https://hanlaboratory.com/ADEC), to facilitate their further investigation.Significance:Our study charts a global landscape of ancestry-associated differences in therapeutic efficacy, highlighting the importance of considering ancestry in anticancer therapies.

A Bayesian adaptive umbrella trial design is proposed to identify effective treatments for biomarker-defined subgroups. This design employs a mixture regression model to jointly model short-term binary and long-term time-to-event outcomes and introduces a latent class model to adaptively cluster subgroups with similar treatment-outcome distributions. The model aims to dynamically borrow information between subgroups in the same class, enhancing the efficiency and robustness in testing subgroup-specific treatment effects. To strengthen the individual ethics of the trial, the design incorporates response-adaptive randomization with early stopping rules for futility and superiority. The simulation results confirm that the proposed design demonstrates desired performance.

The marker-stratified design (MSD) is useful for assessing subgroup-specific treatment effects of molecularly targeted agents (MTA). In MSD, patients are first classified into the marker-positive and marker-negative subgroups, and then randomized for MTA or control treatment within each subgroup. The clinical features of the biomarker and treatments used in MSD offer valuable information for treatment evaluation. Specifically, response rates for patients on the control treatment remain similar across different subgroups if the biomarker involved is not prognostic. Additionally, when the MTA is effective, the marker-positive patients generally exhibit significantly higher response rates compared to the marker-negative patients receiving the same MTA. This paper proposes a Bayesian adaptive design (SSS) that uses these clinical features to enhance the efficiency of MSD. The SSS design employs spike-and-slab priors to dynamically borrow information on response rates across different subgroups. The strength of this information borrowing is automatically determined by two posterior probabilities, which measure the similarities in response rates between different subgroups. Furthermore, an extension of the SSS design is proposed to accommodate patients with missing biomarker profiles, utilizing the Bayesian multiple imputation (BMI) method. Simulation studies confirm that the proposed SSS design demonstrates favorable operational characteristics and outperforms the conventional Bayesian designs.

The win ratio (WR) statistic is increasingly used to evaluate treatment effects based on prioritized composite endpoints, yet existing Bayesian adaptive designs are not directly applicable because the WR is a summary statistic derived from pairwise comparisons and does not correspond to a unique data-generating mechanism. We propose a Bayesian model-assisted adaptive design for randomized phase II clinical trials based on the WR statistic, referred to as the BMW design. The proposed design uses the joint asymptotic distribution of WR test statistics across interim and final analyses to compute posterior probabilities without specifying the underlying outcome distribution. The BMW design allows flexible interim monitoring with early stopping for futility or superiority and is extended to jointly evaluate efficacy and toxicity using a graphical testing procedure that controls the family-wise error rate (FWER). Simulation studies demonstrate that the BMW design maintains valid type I error and FWER control, achieves power comparable to conventional methods, and substantially reduces expected sample size. An R Shiny application is provided to facilitate practical implementation.

Most phase I–II drug‐combination trial designs assume that selecting the optimal dose combination based on early outcomes will also lead to maximum long‐term survival benefits. However, this assumption is often violated in many clinical studies, generally due to high rates of relapse following the initial response. To address this problem, we propose the Great Wall design, a general dose optimization design for drug‐combination trials. The Great Wall design employs a “divide‐and‐conquer” algorithm to address the issue of partial order of toxicity and uses early outcomes to eliminate dose combinations that are excessively toxic or less efficacious. It utilizes a dose randomization approach to construct a candidate set of the promising dose combinations balancing the toxicity and early efficacy outcomes. The patients assigned to the candidate set are followed to collect the survival outcomes and the final optimal dose combination is then selected to maximize the survival benefit. The simulation studies confirm the desirable operating characteristics of the Great Wall design under various clinical settings. R codes are also provided to facilitate the application. The Great Wall design is modular and practically useful in settings where investigators plan to follow patients long enough to assess survival outcomes.

Macrophages integrate microenvironmental cues to orchestrate complex transcriptional and metabolic programs that drive functional polarization. Here, we demonstrate that STK11 links interleukin-4 (IL-4) signaling with metabolic reprogramming to restrain alternatively activated (M2) macrophage polarization. Through integrative transcriptomic and metabolomic analyses, we identified STK11 as a key transcriptional and metabolic regulator during M2 polarization. STK11 deficiency enhanced the expression of M2-associated markers and promoted glutamine metabolism in IL-4–stimulated macrophages. Mechanistically, STK11 deficiency led to increased FOXO1 activation, thereby promoting M2 polarization. Pharmacological inhibition of FOXO1 or glutamine metabolism effectively reversed the enhanced M2 polarization. In an orthotopic model of pancreatic ductal adenocarcinoma, myeloid-specific deletion of STK11 resulted in increased accumulation of M2-like tumor-associated macrophages, impaired antitumor immunity, and accelerated tumor progression. These findings uncover a previously unrecognized role for STK11 in modulating M2 macrophage polarization, offering mechanistic insights that may inform the development of immunometabolic therapies for pancreatic cancer.