Advances in biotechnology have enabled us to assay human tissue and cells to a depth and resolution that was never possible before, redefining what we know as the “biomarker”, and how we define a “disease”. This comes along with the shift of focus from a “one-drug-fits-all” to a “personalized approach”, placing the drug development industry in a highly dynamic landscape, having to navigate such disruptive trends. In response to this, innovative clinical trial designs have been key in realizing biomarker-driven drug development. Regulatory approvals of cancer genome sequencing panels and associated targeted therapies has brought personalized medicines to the clinic. Increasing availability of sophisticated biotechnologies such as next-generation sequencing (NGS) has also led to a massive outflux of real-world genomic data. This review summarizes the current state of biomarker-driven drug development and highlights examples showing the utility and importance of the application of real-world data in the process. We also propose that all stakeholders in drug development should (1) be conscious of and efficiently utilize real-world evidence and (2) re-vamp the way the industry approaches drug development in this era of personalized medicines.

PURPOSE: To understand the mutational landscape of circulating tumor DNA (ctDNA) and tumor tissue of patients with hormone receptor–positive (HR+), human epidermal growth factor receptor-2–negative (HER2–) metastatic breast cancer (MBC) treated with abemaciclib + endocrine therapy (ET). METHODS: Blood samples for ctDNA and/or tissue samples were collected from abemaciclib-treated patients with HR+/HER2– MBC enrolled in the SCRUM-Japan MONSTAR-SCREEN project. Blood samples were collected before abemaciclib initiation (baseline) and at disease progression/abemaciclib discontinuation (post abemaciclib treatment). Clinical and genomic characteristics including neoplastic burden (measured by shedding rate and maximum variant allele frequency [VAF]) were assessed at baseline. Genomic alterations in ctDNA were compared in paired baseline and post abemaciclib treatment samples. RESULTS: All patients (N = 97) were female (median age, 57 years [IQR, 50-67]). In baseline ctDNA (n = 77), PIK3CA (37%), TP53 (28%), ESR1 (16%), and GATA3 (11%) were the most frequently mutated genes. Baseline tissue samples (n = 79) showed similar alteration frequencies. Among patients with baseline ctDNA data, 30% had received previous ET. ESR1 alteration frequency (35% v 8%; P < .01), median shedding rate (3 v 2), and maximum somatic VAF (4 v 0.8; both P < .05) were significantly higher in ctDNA from patients with previous ET than those without previous ET. In paired ctDNA samples (n = 33), PIK3CA and ESR1 alteration frequencies were higher after abemaciclib treatment than at baseline, though not statistically significant. Among the post-treatment alterations, those newly acquired were detected most frequently in FGF3/4/19 (18%); PIK3CA, TP53, CCND1, and RB1 (all 15%); and ESR1 (12%). CONCLUSION: We summarized the ctDNA and cancer tissue mutational landscape, including overall neoplastic burden and PIK3CA and ESR1 hotspot mutations in abemaciclib-treated patients with HR+/HER2– MBC. The data provide insights that could help optimize treatment strategies in this population.

Nucleic acid (NA) biomarkers play critical roles in drug development. However, the global regulatory guidelines for assessing quantification methods specific to NA biomarkers are limited. The valid...

Introduction: Atopic dermatitis (AD) is an inflammatory skin disease that is heterogeneous in clinical presentation and biological mechanisms. Several studies have suggested biomarker-defined molecular endotypes in AD. This study aimed to characterize potential endotypes in Japanese patients with moderate-to-severe AD and comprehensively evaluate their circulating protein profiles to better understand disease etiology. Methods: Serum samples from Japanese patients with moderate-to-severe AD ( n = 73) enrolled in a phase 3 study of baricitinib (BREEZE-AD2; NCT03334422) and samples from healthy controls ( n = 15) were analyzed using the Olink Explore 1536 assay. Patient clusters were identified through k -means clustering. Differential expression analysis and weighted gene co-expression network analysis were performed for in-depth examination of proteomic profiles. Results: Two patient clusters, characterized by high (AD_HI) and low (AD_LO) inflammatory profiles, were found to be stable and reproducible. Canonical AD inflammatory mediators—including interleukin (IL)-13, IL-19, pulmonary and activation-regulated chemokine (PARC), thymus and activation-regulated chemokine (TARC), chemokine (C-C motif) ligand (CCL)22, CCL26, and CCL27—were upregulated in both clusters, with greater upregulation in the AD_HI cluster. Additionally, proteins not typically associated with AD-related inflammation were upregulated in AD_HI patients. The AD_HI cluster was associated with protein networks representing a range of immune and non-immune pathways. Dysregulated protein signatures associated with the AD_HI cluster were also correlated with skin-based disease severity scores. Conclusion: This study characterizes the circulating proteome and clinical characteristics across putative molecular endotypes in AD. The findings corroborate current knowledge on AD pathophysiology and suggest other axes of dysregulation in a subset of patients with AD. These results may support the development of personalized therapeutic approaches.