Category: 3. Business

Introduction

Diabetic retinopathy (DR), a primary complication of diabetes mellitus, is a leading cause of visual impairment in the working-age population worldwide, including China.¹ According to the microangiopathy caused by hyperglycemia, DR can be divided into two stages: nonproliferative DR (NPDR) and proliferative DR (PDR). Chronic hyperglycemia causes retinal vasodilation and hemodynamic changes, damage to endothelial cells and pericytes, and thickening of the vascular basement membrane, ultimately forming retinal microaneurysms and punctate retinal hemorrhage.² As angiopathy develops, retinal capillaries are not perfused, and retinal hypoxia induces neovascularization, ultimately leading to irreversible vision loss.^3,4

Current therapeutic strategies for DR primarily target modifiable risk factors, with systemic management including stringent glycemic control, blood pressure regulation, and lipid-lowering therapy demonstrating clinical efficacy.⁵ However, the precise molecular mechanisms underlying metabolic dysregulation-induced retinal dysfunction remain unclear.⁶ Emerging evidence suggests that DR shares inflammatory pathways with systemic diabetic complications, with immune dysregulation being increasingly recognized as a pivotal factor that disrupts retinal immune homeostasis.³ Sustained, low-grade inflammation resulting from immune dysregulation plays a pivotal role in the pathogenesis of DR-related retinal vascular damage. Key manifestations include blood–retinal barrier disruption, macular edema, and pathological neovascularization, all of which contribute to vision-threatening complications.⁷ Machine learning approaches using clinical and glycemic data may effectively predict DR outcomes.⁸ The formation and maintenance of the retinal immune balance depend on many factors, among which the inhibitory cell population and cytokines are important regulatory factors. Therefore, investigating the role of immune cells in retinal pathology provides a new avenue for the development of effective therapeutic strategies.

Myeloid-derived suppressor cells (MDSCs), a heterogeneous population of immunoregulatory cells, play critical roles in maintaining immune homeostasis through their immunosuppressive functions.^9,10 Extensive research has established associations between MDSC populations and various pathological conditions, with peripheral blood MDSC expansion frequently observed in malignancies, infectious diseases, and chronic inflammatory states. MDSCs exhibit unique functional properties, such as the production of reactive oxygen species, peroxidase, nitric oxide, and anti-inflammatory cytokines, along with angiogenesis. MDSCs have been identified in various non-tumor pathologies, including autoimmune disorders and chronic inflammatory diseases.¹¹ Clinical observations have demonstrated MDSC expansion in patients with type 1 and type 2 diabetes mellitus (T2DM).^12,13 A high proportion of MDSCs in retinal tissue can promote retinal vascular inflammation and angiogenesis, which is an important pathological process in DR.^13,14 MDSCs are primarily categorized into two subsets: monocytic (M-MDSCs) and granulocytic (G-MDSCs). This study aimed to quantify peripheral blood M-MDSC and G-MDSC levels in patients with DR to investigate their potential involvement in DR pathogenesis.

Materials and Methods

Ethics Statement

This study received ethical approval from the Institutional Review Board at the Affiliated Changshu Hospital of Nantong University (Approval No. 2023-KY-SKQ-01) and adhered to the ethical guidelines outlined in the Declaration of Helsinki, including risk–benefit assessment and participant confidentiality protection. Informed consent was obtained from all participants, and a detailed explanation of the study purpose and voluntary withdrawal rights was provided before blood sample collection.

Sample Size Calculation for the Study

This study was designed as a three-arm, parallel-group trial (NPDR group, PDR group, and healthy control group), with peripheral blood MDSC levels (%) as the primary outcome measure. Based on preliminary experimental data, sample size calculation was performed using the PASS (Power Analysis and Sample Size, version 15.0; NCSS, LLC, Kaysville, UT, USA). A one-way ANOVA model was selected with the following parameters: α = 0.05 (adjusted to α = 0.0167 after Bonferroni correction for multiple comparisons), power = 80%, and effect size (Cohen’s f) = 0.52 (calculated from the group means and pooled standard deviation). The PASS analysis indicated that 19 participants per group would be required.

Patients

From January 2023 to February 2024, 42 patients with T2DM and DR and 20 age- and sex-matched healthy controls were enrolled at the Affiliated Changshu Hospital of Nantong University (Changshu, China). The DR cohort comprised 21 patients with PDR and 21 with NPDR. T2DM diagnosis was confirmed by endocrinologists according to the American Diabetes Association diagnostic criteria. DR staging was performed by ophthalmologists through comprehensive fundus examinations including fundus photography and optical coherence tomography, with severity classified according to the International Clinical Diabetic Retinopathy Severity Scale.

The exclusion criteria were as follows:

1. Diagnosis of type 1 diabetes mellitus or other specific diabetes subtypes;
2. History of autoimmune disorders, neurodegenerative diseases, or malignancies;
3. Previous major cardiovascular/cerebrovascular events (eg, myocardial infarction, ischemic stroke, or cerebral hemorrhage);
4. Active systemic infections or inflammatory conditions;
5. Use of glucocorticoids or immunosuppressive agents within the past 3 months.

Data and Sample Collection

Demographic and clinical characteristics were systematically collected from all participants. Whole blood samples were obtained in ethylenediaminetetraacetic acid (EDTA)-treated blood collection tubes and processed within 2 h for subsequent analysis of MDSCs via flow cytometry.

Flow Cytometric Analysis

Peripheral blood mononuclear cells (PBMCs) were isolated from anticoagulated whole blood via density gradient centrifugation (Ficoll-Paque™ PLUS density gradient media, Cytiva, Uppsala, Sweden). Cells were washed twice with phosphate-buffered saline (PBS) and resuspended in 100 μL PBS for staining. Samples were incubated with fluorescently labeled antibodies targeting CD45, Lin, CD56, HLA-DR, CD16, CD11b, CD33, CD14, and CD15 (BioLegend, San Diego, CA, USA). Antibodies were used at manufacturer-recommended concentrations, with a volume of 1 μL per antibody per 100 μL. PBMCs were initially gated based on side and forward scatter properties. Thereafter, dead cells were excluded, and the CD45⁺ viable cell population was gated. The Lin-negative population was identified using Lin/FSC gating, and the CD56⁻CD16⁻HLA⁻DR⁻ population was delineated using logical gating. The MDSC-specific antibodies, CD33 and CD11b were then used to isolate the CD33⁺CD11b⁺ population from the previously identified negative population, which was further classified into CD15⁺ G-MDSCs and CD14⁺ M-MDSCs. Fluorescence minus one (FMO) controls (CD15⁻FMO or CD14⁻FMO) were used. In addition, the blank control group was not treated with reagents. Flow cytometry data were acquired using a Beckman Coulter Dxflex flow cytometer (Beckman Coulter, Inc., CA, USA) and analyzed with the FlowJo software (version 10.6.2; Tree Star Inc., OR, USA). All flow cytometric analyses were conducted by a single experienced technician (Z.Z.) to maintain consistency and reproducibility. Stringent quality control protocols were implemented, including regular assessment of optical alignment, fluidics stability, and fluorescence sensitivity to confirm adherence to manufacturer specifications.

Statistical Analysis

Data are expressed as the mean ± standard deviation for continuous variables or as counts and percentages for categorical variables. The normality of quantitative data was assessed using histograms and the Shapiro–Wilk test. Intergroup comparisons were performed using Student’s t-test or Wilcoxon rank-sum test for normally and non-normally distributed data, respectively. Multiple-group comparisons were performed using a one-way analysis of variance or Kruskal–Wallis test, depending on data distribution. Categorical variables were analyzed using the Pearson chi-square test, with post hoc pairwise comparisons adjusted using the Bonferroni method. Correlations between variables were evaluated using Spearman’s rank correlation coefficient. All statistical analyses were performed using SPSS 25.0 (IBM Corp., Armonk, NY, USA). GraphPad Prism 8.0 (GraphPad Software Inc., San Diego, CA, USA) was used to generate a histogram. Statistical significance was set at a two-sided P-value < 0.05.

Results

Baseline Characteristics

This study included 42 patients with DR, 21 of whom had NPDR. The demographic characteristics of the participants are represented in Table 1. Sex and age distributions did not significantly differ among the NPDR, PDR, and healthy control groups (P = 0.811 and P = 0.163, respectively).

Table 1 Baseline Characteristics of the Participants

Proportion of Circulating MDSCs in the Three Groups

The MDSCs in the peripheral blood were analyzed via flow cytometry (Figure 1A). The difference in the percentage of peripheral blood M-MDSCs significantly differed among the three groups (Table 2). Multiple comparisons revealed significant differences among the PDR, NPDR, and healthy control groups (P = 0.0063 and P = 0.0056; Figure 1B). No significant differences were observed between the NPDR and healthy control groups. Meanwhile, the percentage of peripheral blood G-MDSCs did not significantly differ among the three groups (P = 0.226; Figure 1C).

Table 2 Proportion of Circulating MDSCs in the Three Groups

Figure 1 Fluorescence-activated cell sorting analysis of circulating myeloid-derived suppressor cells (MDSCs) in the blood of the participants. Peripheral blood mononuclear cells were initially gated based on side and forward scatter properties. Thereafter, dead cells were excluded, and the CD45+ viable cell population was gated. The Lin-negative population was identified using Lin/FSC gating. Subsequently, the CD56−CD16−HLA−DR− population was delineated through logical gating. The MDSC-specific antibodies CD33 and CD11b, were then used to isolate the CD33+CD11b+ population from the previously identified negative population, which was further classified into CD15+G-MDSCs and CD14+M-MDSCs. (A) Flow cytometric analysis. (B) Circulating M-MDSC ratios in the nonproliferative diabetic retinopathy (NPDR), proliferative DR (PDR), and healthy control groups. (C) Circulating G-MDSC ratios in the NPDR, PDR, and healthy control groups.

Correlations Among the Proportions of Circulating M-MDSC Levels in Patients with DR

The proportion of M-MDSCs in the peripheral blood of patients positively correlated with fasting blood glucose (r = 0.282; P = 0.027) and glycosylated hemoglobin levels (r = 0.329; P = 0.009) (Table 3).

Table 3 Correlation Between Clinical Characteristics and Circulating M-MDSC Levels

Discussion

We analyzed the proportions of M-MDSCs and G-MDSCs in the peripheral blood of patients with different stages of DR and their correlation with other factors. The proportion of M-MDSCs in peripheral blood significantly increased in the PDR group and positively correlated with fasting blood glucose and glycosylated hemoglobin levels.

MDSCs are defined as heterogeneous myeloid cell groups at different stages of differentiation. In bone marrow and secondary lymphoid organs, changes or interruptions in the generation and maturation processes of bone marrow progenitor cells can produce MDSCs. The reprogramming of mature myeloid cells in peripheral tissues can produce MDSCs. Under pathological conditions, heterogeneous myeloid cells exhibit immunosuppressive activity, a characteristic observed in cancer as well as in chronic inflammation, infection, autoimmunity, and allergy.¹⁵ M-MDSCs are widely recognized for their key role in angiogenesis in tumors. They promote neovascularization and establish and maintain the tumor vascular system by secreting vasoactive substances, cytokines, and chemokines.^16,17 The levels of M-MDSCs were elevated in the peripheral blood of patients with PDR, indicating their potential role in retinal neovascularization.

Previous studies have reported an increased proportion of M-MDSCs in the peripheral blood of patients with type 2 diabetes, aligning with the findings of the current study.^18,19 Our results further revealed that the levels of M-MDSCs significantly increased in patients with PDR, who had worse glycemic control indices than in those with NPDR. Elevated M-MDSC levels also positively correlated with glycemic control indices. These findings further reveal the core role of M-MDSCs in PDR progression.

The lack of significant differences in G-MDSC levels across study groups may reflect their distinct biological characteristics. As G-MDSCs primarily consist of immature neutrophils that typically expand during acute inflammatory responses or in immunosuppressive tumor microenvironments,¹⁵ their dynamics may differ in the context of chronic, low-grade inflammation characteristic of DR. Additionally, the limited sample size may have reduced our power to detect subtle variations in this heterogeneous cell population.

The pathological progression from NPDR to PDR involves complex vascular changes driven by chronic hyperglycemia, including retinal ischemia, increased vascular permeability, and endothelial dysfunction.⁶ The elevated M-MDSC levels observed in patients with PDR may contribute to this progression through multiple mechanisms. MDSCs play a pivotal role in promoting tumor angiogenesis through multiple mechanisms. These cells secrete a diverse array of proangiogenic factors, including matrix metalloproteinase-9 (MMP-9), vascular endothelial growth factor-A, fibroblast growth factor-2 (FGF-2), platelet-derived growth factor, interleukin-1β (IL-1β), IL-28, transforming growth factor β, epidermal growth factor, and hepatocyte growth factor. Granulocyte colony-stimulating factor enhances myeloid-dependent angiogenesis by stimulating MDSCs to express prokineticin Bv8 through the signal transducer and activator of transcription 3 (STAT3) signaling pathway. The STAT3 signaling in M-MDSCs stimulates the production of proangiogenic factors, including VEGF and FGF-2. Additionally, STAT3-activating factors released from these M-MDSCs can induce STAT3 signaling in endothelial cells, further amplifying neoangiogenesis through the upregulation of VEGF receptor 2 and MMPs.^15,20 Furthermore, MDSCs can directly differentiate into endothelial cells, thereby contributing to vascular mimicry formation and establishing an alternative angiogenesis pathway independent of conventional endothelial cell-mediated vessel formation.¹⁷ Supporting evidence comes from studies on retinopathy of prematurity (ROP), where flow cytometry revealed significantly higher circulating MDSC levels in 14 infants with ROP compared to controls. In oxygen-induced retinopathy models, MDSCs were markedly elevated during the neovascular phase and preferentially accumulated in pathological lesions. MDSC depletion attenuated retinal vascular occlusion and pathological angiogenesis, suggesting their crucial role in promoting aberrant vessel formation. MDSCs may modulate the inflammatory microenvironment through secretion of pro-inflammatory factors (CCL2, IL-1β, TNF-α), thereby influencing angiogenic processes. While MDSCs regulate microglial recruitment to neovascular areas, their depletion does not affect microglial activation status. Transcriptomic analyses further implicate MDSCs in pathological pathways involving inflammation, immunity, and neuroinflammation. Similarly, in PDR, M-MDSCs may facilitate retinal neovascularization by upregulating proangiogenic factors such as VEGF and MMP-9.^21,22 The involvement of MMP9 in PDR pathogenesis is well-documented, and its secretion by M-MDSCs could enhance VEGF bioavailability, further driving pathological angiogenesis.^23,24

In patients with diabetes, retinal pigment epithelial, glial, and endothelial cells regulate the production of monocyte chemoattractant protein-1 (MCP-1), which exerts its function through its receptor, C-C chemokine receptor type 2 (CCR2). Monocytes and macrophages express CCR2 and are involved in inflammatory events in DR. In patients with DR, MCP-1 levels in the vitreous humor are higher than those in serum, suggesting that MCP-1 is locally expressed and produced in the eye. In addition, the expression level of MCP-1 in the vitreous humor positively correlates with the severity of DR.²⁵ Therefore, MCP-1 likely drives pathological retinal neovascularization by recruiting M-MDSCs to the retina, which exacerbates vascular leakage and amplifies proangiogenic signaling via VEGF upregulation.

Previous transcriptome sequencing studies have reported that M-MDSCs may regulate retinal neoangiogenesis through inflammation-related pathways.¹⁴ Therefore, M-MDSCs may play a role in regulating retinal neovascularization through pathways related to inflammation and immunity.

Despite the findings, this study has certain limitations. The relatively small sample size may limit the generalizability of the results, underscoring the need for larger-scale, multicenter studies to validate these observations. Future investigations could focus on elucidating the temporal dynamics of M-MDSC levels throughout DR progression, emphasizing their potential as predictive biomarkers for the transition from NPDR to PDR. Additionally, comprehensive analyses accounting for confounding factors such as medication use and diabetes duration are warranted. Further mechanistic studies are essential to unravel the regulatory pathways of MDSCs and evaluate their therapeutic potential as novel treatment targets for DR.

Conclusion

This study demonstrated a significant elevation in the proportion of circulating M-MDSCs in patients with PDR, indicating their potential role in the pathogenesis and progression of PDR. Mechanistically, this phenomenon may be closely linked to the proangiogenic effects of retinal M-MDSCs. Furthermore, a strong positive correlation was observed between circulating M-MDSC levels and glycemic control markers, including fasting blood glucose and HbA1c. These findings provide novel insights into the immune mechanisms underlying PDR and underscore the importance of glycemic control in the prevention and management of DR.

Data Sharing Statement

Study data are available from the corresponding author upon reasonable request to protect participant privacy.

Consent for Publication

All participants provided written general broad consent upon hospital admission, which explicitly authorized the use of their anonymized medical data for future research and scientific publication. This study utilized retrospectively collected clinical data under the approval of the Ethics Committee of the Affiliated Changshu Hospital of Nantong University (Approval No. 2023-KY-SKQ-01), with waived requirement for additional study-specific consent.

Author Contributions

All authors contributed to data analysis, drafting or revising the article, have agreed on the journal to which the article will be submitted, gave final approval of the version to be published, and agree to be accountable for all aspects of the work.

Funding

This study was funded by the Suzhou Science and Technology Development Plan Project (SKY2023088), Key Laboratory Open Topic Funding Project in Jiangsu Universities (XZSYSKF2023020), Jiangsu Provincial Natural Science Foundation Project (BK20231201), Open Project of State Key Laboratory of Radiation Medicine of Soochow University (GZK1202305), Nantong University Clinical Medicine Special Research Fund Project (2024JY029), and Jiangsu University Medical Education Collaborative Innovation Funding (JDYY2023099).

Disclosure

The authors declare that there is no conflict of interest regarding the publication of this paper.

References

1. Steinmetz JD, Bourne RRA, Briant PS. GBD 2019 Blindness and Vision Impairment Collaborators, Vision Loss Expert Group of the Global Burden of Disease Study. Causes of blindness and vision impairment in 2020 and trends over 30 years, and prevalence of avoidable blindness in relation to VISION 2020: the right to sight: an analysis for the global burden of disease study. Lancet Glob Health. 2021;9(2):e144–60. doi:10.1016/S2214-109X(20)30489-7

2. Sheng X, Zhang C, Zhao J, et al. Microvascular destabilization and intricated network of the cytokines in diabetic retinopathy: from the perspective of cellular and molecular components. Cell Biosci. 2024;14(1):85. doi:10.1186/s13578-024-01269-7

3. Pan WW, Lin F, Fort PE. The innate immune system in diabetic retinopathy. Prog Retin Eye Res. 2021;84:100940. doi:10.1016/j.preteyeres.2021.100940

4. Lei Y, Wang Y, Tang S, Yang J, Lai D, Qiu Q. The adaptive immune system in the retina of diabetics. Surv Ophthalmol. 2025;70(2):241–254. doi:10.1016/j.survophthal.2024.11.005

5. Cheung N, Mitchell P, Wong TY. Diabetic retinopathy. Lancet. 2010;376(9735):124–136. doi:10.1016/S0140-6736(09)62124-3

6. Antonetti DA, Silva PS, Stitt AW. Current understanding of the molecular and cellular pathology of diabetic retinopathy. Nat Rev Endocrinol. 2021;17(4):195–206. doi:10.1038/s41574-020-00451-4

7. Mesquida M, Drawnel F, Fauser S. The role of inflammation in diabetic eye disease. Semin Immunopathol. 2019;41(4):427–445. doi:10.1007/s00281-019-00750-7

8. Montaser E, Shah VN. Prediction of incident diabetic retinopathy in adults with type 1 diabetes using machine learning approach: an exploratory study. J Diabetes Sci Technol. 2024;19322968241292369.

9. Bronte V, Brandau S, Chen S-H, et al. Recommendations for myeloid-derived suppressor cell nomenclature and characterization standards. Nat Commun. 2016;7(1):12150. doi:10.1038/ncomms12150

10. Zhou J, Zhang M, Ju X, et al. Increased monocytic myeloid-derived suppressor cells in type 2 diabetes correlate with hyperglycemic and was a risk factor of infection and tumor occurrence. Sci Rep. 2024;14(1):4384. doi:10.1038/s41598-024-54496-w

11. Sendo S, Saegusa J, Morinobu A. Myeloid-derived suppressor cells in non-neoplastic inflamed organs. Inflamm Regen. 2018;38(1):19. doi:10.1186/s41232-018-0076-7

12. F W-L, F J, B J, Ma S. Myeloid-derived suppressor cells are increased in frequency but not maximally suppressive in peripheral blood of type 1 diabetes mellitus patients. Clin Immunol. 2014;153.

13. Ghemiș L, Goriuc A, Minea B, et al. Myeloid-derived suppressor cells (MDSCs) and obesity-induced inflammation in type 2 diabetes. Diagnostics. 2024;14(21):2453. doi:10.3390/diagnostics14212453

14. Wu X, Zhong L, Yu J, et al. MDSCs promote pathological angiogenesis in ocular neovascular disease. Biomed Pharmacother. 2024;178:117222. doi:10.1016/j.biopha.2024.117222

15. Lasser SA, Ozbay Kurt FG, Arkhypov I, Utikal J, Umansky V. Myeloid-derived suppressor cells in cancer and cancer therapy. Nat Rev Clin Oncol. 2024;21(2):147–164. doi:10.1038/s41571-023-00846-y

16. Zalpoor H, Aziziyan F, Liaghat M, et al. The roles of metabolic profiles and intracellular signaling pathways of tumor microenvironment cells in angiogenesis of solid tumors. Cell Commun Signal. 2022;20(1):186. doi:10.1186/s12964-022-00951-y

17. Yang F, Lee G, Fan Y. Navigating tumor angiogenesis: therapeutic perspectives and myeloid cell regulation mechanism. Angiogenesis. 2024;27(3):333–349. doi:10.1007/s10456-024-09913-z

18. Fernández-Ruiz JC, Galindo-De Ávila JC, Martínez-Fierro ML, et al. Myeloid-derived suppressor cells show different frequencies in diabetics and subjects with arterial hypertension. J Diabetes Res. 2019;2019:1568457. doi:10.1155/2019/1568457

19. Wang S, Tan Q, Hou Y, Dou H. Emerging roles of myeloid-derived suppressor cells in diabetes. Front Pharmacol. 2021;12:798320. doi:10.3389/fphar.2021.798320

20. Kujawski M, Kortylewski M, Lee H, Herrmann A, Kay H, Yu H. Stat3 mediates myeloid cell-dependent tumor angiogenesis in mice. J Clin Invest. 2008;118(10):3367–3377. doi:10.1172/JCI35213

21. Du R, Lu KV, Petritsch C, et al. HIF1alpha induces the recruitment of bone marrow-derived vascular modulatory cells to regulate tumor angiogenesis and invasion. Cancer Cell. 2008;13(3):206–220. doi:10.1016/j.ccr.2008.01.034

22. Ahn G-O, Brown JM. Matrix metalloproteinase-9 is required for tumor vasculogenesis but not for angiogenesis: role of bone marrow-derived myelomonocytic cells. Cancer Cell. 2008;13(3):193–205. doi:10.1016/j.ccr.2007.11.032

23. Singh K, Goyal P, Singh M, et al. Association of functional SNP-1562C > T in MMP9 promoter with proliferative diabetic retinopathy in north Indian type 2 diabetes mellitus patients. J Diabetes Complications. 2017;31(12):1648–1651. doi:10.1016/j.jdiacomp.2017.08.010

24. Jayashree K, Yasir M, Senthilkumar GP, Ramesh Babu K, Mehalingam V, Mohanraj PS. Circulating matrix modulators (MMP-9 and TIMP-1) and their association with severity of diabetic retinopathy. Diabetes Metab Syndr. 2018;12(6):869–873. doi:10.1016/j.dsx.2018.05.006

25. Taghavi Y, Hassanshahi G, Kounis NG, Koniari I, Khorramdelazad H. Monocyte chemoattractant protein-1 (MCP-1/CCL2) in diabetic retinopathy: latest evidence and clinical considerations. J Cell Commun Signal. 2019;13(4):451–462. doi:10.1007/s12079-018-00500-8

The first half of 2025 has seen a surge of FDA approvals that expand treatment options and improve access to care for patients across multiple pediatric and adolescent health conditions.

From novel vaccines and needle-free emergency treatments to groundbreaking therapies for rare diseases, HIV prevention, and autoimmune disorders, these approvals reflect critical strides in both innovation and accessibility. Many of the newly authorized therapies offer more convenient dosing, at-home administration, or first-in-class mechanisms, addressing long-standing clinical gaps and supporting personalized care.

In this midyear roundup, Contemporary Pediatrics highlights the most impactful FDA decisions from January through June 2025 that are shaping the future of care for infants, children, and teens.

Click on each title below to read more.

1. FDA approves GSK’s meningococcal ABCWY vaccine

On February 14, 2025, the FDA approved the meningococcal ABCWY (MenABCWY, Penmenvy; GSK) vaccine for individuals aged 10 to 25 years, offering protection against the five major serogroups (A, B, C, W, Y) of Neisseria meningitidis, which cause invasive meningococcal disease (IMD).

The vaccine combines components of the meningococcal group B vaccine (Bexsero) and the ACYW conjugate vaccine (Menveo). Approval was based on a phase 3 trial (NCT04502693) involving 3,650 participants across seven countries, demonstrating strong immunogenicity, non-inferiority to existing vaccines, and a well-tolerated safety profile. MenB accounts for over half of IMD cases in Sadolescents and young adults, yet vaccination rates remain low.

2. FDA approves neffy 1 mg for anaphylaxis in children aged 4 years and older

The FDA approved epinephrine nasal spray (neffy; ARS Pharmaceuticals) 1 mg on March 5, 2025, for children aged 4 years and older weighing 33 to 66 lbs, making it the first needle-free epinephrine treatment for younger pediatric patients. This expands its indication beyond the original August 2024 approval for patients weighing at least 66 lbs.

Expected to launch in the United States by May 2025, neffy offers a user-friendly, temperature-stable alternative to traditional epinephrine auto-injectors, addressing fears of needle-based administration that often delay treatment. Approval was based on pharmacokinetic and pharmacodynamic data showing comparable efficacy to injectable epinephrine.

3. FDA approves diazoxide choline extended-release tablets for hyperphagia in Prader-Willi syndrome

The FDA approved diazoxide choline extended-release tablets (VYKAT XR; Soleno Therapeutics) on March 26, 2025, for treating hyperphagia in individuals aged 4 years and older with Prader-Willi syndrome (PWS). Granted priority review, breakthrough therapy, fast track, and orphan drug designations, VYKAT XR is the first approved treatment specifically addressing hyperphagia, a hallmark symptom of PWS.

In phase 3 trials, the treatment significantly improved hyperphagia, aggressive behaviors, fat mass, and metabolic parameters, with patients showing greater reductions in hyperphagia scores at 26 and 52 weeks compared to natural history data (P < 0.001). PWS, which affects approximately 1 in 15,000 live births, is associated with persistent hunger, cognitive disabilities, and metabolic challenges. Advocacy groups and researchers hailed the approval as a critical advancement for the PWS community.

4. FDA approves clesrovimab to protect infants during first RSV season

On June 9, 2025, the FDA approved Merck’s clesrovimab (Enflonsia), an extended half-life monoclonal antibody, to protect infants from respiratory syncytial virus (RSV) during their first RSV season. The approval was supported by positive results from the phase 2b/3 CLEVER and phase 3 SMART trials, which demonstrated that a single dose of clesrovimab significantly reduced RSV-related infections and hospitalizations in both healthy and high-risk infants. In the CLEVER trial, clesrovimab reduced medically attended lower respiratory infections by 60.5%, RSV-related hospitalizations by 84.3%, and severe RSV cases by over 90% through 5 months post-dose (P < 0.001 for all). Unlike other monoclonal antibodies, clesrovimab targets site 4 of the RSV fusion protein, neutralizing more than 96% of RSV A and B strains. The therapy, given as a fixed intramuscular dose, offers direct, rapid, and durable protection throughout a typical RSV season.

5. FDA approves new presentation of ustekinumab-stba biosimilar for plaque psoriasis, psoriatic arthritis

On June 15, 2025, the FDA approved a new presentation of ustekinumab-stba (Steqeyma; Celltrion), a biosimilar to ustekinumab (Stelara), in a 45mg/0.5mL single-dose vial for subcutaneous injection. The new dosage form is indicated for patients aged 6 to 17 years weighing under 60 kg with plaque psoriasis or psoriatic arthritis. With this approval, ustekinumab-stba matched all dosing forms of the reference product. The decision was supported by phase 3 data showing clinical similarity to the reference biologic, with no meaningful differences in safety or efficacy.

6. FDA approves garadacimab-gxii to prevent HAE attacks in patients aged 12 years, older

On June 17, 2025, FDA approved CSL’s garadacimab-gxii (Andembry), a first-in-class monoclonal antibody targeting factor XIIa, to prevent hereditary angioedema (HAE) attacks in patients aged 12 years and older. Approval was based on the phase 3 VANGUARD trial, where once-monthly garadacimab-gxii reduced mean HAE attack rates by 86.5% and median attack rates by over 99% compared to placebo (P < 0.001). An open-label extension study also showed sustained efficacy and a favorable long-term safety profile. Experts highlighted its novel mechanism and convenient dosing as key advances for HAE management. Meanwhile, FDA review of another HAE therapy, sebetralstat, has been delayed due to resource constraints, though no concerns about its safety or efficacy were raised.

7. FDA approves twice-yearly lenacapavir as PrEP for HIV in adolescents, adults

On June 18, 2025, the FDA approved lenacapavir (Yeztugo; Gilead Sciences), the first and only twice-yearly injectable HIV-1 capsid inhibitor for pre-exposure prophylaxis (PrEP) to reduce the risk of sexually acquired HIV in individuals aged 12 and older weighing at least 35 kg. Approval was based on results from the phase 3 PURPOSE 1 and PURPOSE 2 trials, in which lenacapavir showed no HIV infections among cisgender women in sub-Saharan Africa (PURPOSE 1) and just 2 infections among 2179 participants in a diverse population including cisgender men and gender-diverse people (PURPOSE 2), achieving over 99.9% effectiveness in both trials. The treatment demonstrated superiority over once-daily oral Truvada (F/TDF), offering a long-acting option for people who prefer or require less frequent dosing. Experts hailed the approval as a major step forward in HIV prevention, potentially improving adherence and reducing stigma associated with daily PrEP.

8. FDA approves at-home belimumab autoinjector for children with lupus nephritis

On June 24, 2025, the FDA approved a 200 mg/mL autoinjector formulation of belimumab (Benlysta; GSK) for at-home use in children aged 5 years and older with active lupus nephritis (LN), making it the first subcutaneous, self-administered biologic treatment for pediatric LN. Originally approved for adult SLE in 2011 and expanded to pediatric indications in later years, this new formulation offers a convenient alternative to intravenous infusions, potentially reducing clinic visits and easing treatment burdens for families. The decision was based on belimumab’s established safety and efficacy profile. Experts noted that pediatric lupus is often more severe than adult-onset disease, making flexible treatment options critical. The autoinjector is now available for eligible patients and can be administered by a health care provider or trained caregiver at home.

Story Continues

Dynamic adaptation of A(H3N2) virus in mice

To obtain an A(H3N2)-MA strain, we conducted successive lung-to-lung passages of the A(H3N2)-WT strain in C57BL/6J mice. The A(H3N2) viruses identified in 17 passages were designated as MA1–MA17, respectively (Fig. 1A). According to the changes in body weight, survival state, lung index (a ratio of the weight of the lung to the body weight of the mouse), and lung viral RNA loads of the infected mice, MA1, MA5, MA8, MA13, and MA17 were selected as key passages in the following analysis.

In the MA1, no body weight loss was observed in mice infected with the A(H3N2)-WT strain (Fig. 1B), and all mice survived during the observation period (Fig. 1C). The lung indexes of the infected mice showed no noticeable changes compared to the controls (Fig. 1D), and virus was detected in one of the three mice at 5 dpi. In the MA5, the infected mice have a decrease of approximately 5% in body weight at 7 dpi and later the body weight returned to a normal state. No mice died, but the lung indexes and lung virus RNA loads of the infected mice were higher than those of the control group (Fig. 1B–E).

As the number of passages increased, the body weight losses of the infected mice increased compared with their initial body weights before virus infection. Specifically, from 5 to 7 dpi, the rates of body weight reduction were greater than 22% for MA8, 25% for MA13, and 25% for MA17 compared to their initial body weights (Fig. 1B). In addition, in the MA8 passage, the survival rate of the infected mice dropped to 33%, while all mice in the MA13 and MA17 passages died within 7 and 5 dpi, respectively (Fig. 1C). Additionally, the lung indexes of the infected mice gradually increased as passage increased. The lung indexes for mice from the MA13 and MA17 passages were significantly higher than that of the control group, respectively (Fig. 1D). Meanwhile, the lung virus RNA loads of mice from the MA5, MA8 and MA13 passages progressively increased. The mean virus RNA loads of MA17 was comparable with that of the MA13 passage, which were both significantly higher than that of MA1 (Fig. 1E). These results suggest that as the passage number increased, the A(H3N2) virus from the infected mice showed gradual enhancement in their virulence, replication ability, and adaptation to mice.

Adaptation of A(H3N2) in mice involves dynamic mutations of the virus

To reveal the molecular basis correlated to the adaptation of the A(H3N2) virus in mice, we analyzed AA mutations that occurred in the viruses from five key passages (MA1, MA5, MA8, MA13, and MA17). This analysis was based on the genomic data generated using NGS method. We identified fourteen mutations that are unique in the A(H3N2)-MA virus rather than the A(H3N2)-WT strain (Table 1). These AA mutations include twelve nonsynonymous mutations (PB2-S590R, PB1-I682V, PA-K615E, PA-X-F246S, HA-N91T, HA-N122D, HA-K207E, HA-I242T, HA-N246K, HA-M478I, NP-G384R, and M1-D232N) in PB2, PB1, PA, PA-X, HA, NP, and M1 genes, and two synonymous mutations (PA-G(GGG)462G(GGA) and PA-L(CTT)246L(CTC)) in PA gene.

Among these, four mutations are prone to occur during the mice infection. The HA-N246K (H3 numbering throughout), HA-M478I, NP-G384R, and M1-D232N mutations occurred in the MA1 passage, with a higher proportion of greater than 90%. In addition, six of the fourteen mutations (i.e., PB2-S590R, PB1-I682V, PA-K615E, PA-X-F246S,PA-G(GGG)462G(GGA), and PA-L246L) have a gradual accumulation progress during the passaging. The PB2-S590R, PA-K615E, and PA-G(GGG)462G(GGA) mutations occurred in the MA1 passage, with proportions of 1.47%, 1.27%, and 7.56%, respectively, and their proportions reached 97.73%, 98.57%, and 98.59% in the MA17 passage, respectively. The PB1-I682V, PA-X-F246S, and PA-L(CTT)246L(CTC) mutations were first found in the MA5, MA8, and MA8, respectively. Furthermore, three mutations on the HA genes occurred at the late stage of the virus adaptation, and the HA-N91T, HA-K207E, and HA-I242T mutations were first observed in the MA13 passage, with proportions of 11.6%, 11.16%, and 11.18%, respectively; and these mutations increased to approximately 70% in the MA17 passage. In addition, the HA-N122D had already occurred in the MA5 passage with a proportion of 97.75% but was not found in the MA1 passage.

Moreover, eight mutations, PB2-S590R, PA-K615E, HA-N122D, HA-N246K, HA-M478I, NP-G384R, M1-D232N, and PA-G(GGG)462G(GGA) with the proportion greater than 95% were kept in the MA17 passage. In addition, the other mutations took an occupation of greater than 50% in the MA17 passage.

Mouse adapted mutation enhances the polymerase activity of the RNP complex

To investigate the impact of mouse-adapted mutations in the PB2, PB1, PA, and NP genes on polymerase activity, we analyzed the activity of the RNP complex from these gene combinations. As shown in Fig. 2, the relative polymerase activity of the RNP complex containing PB2-S590R, PB1-I682V, PA-K615E, or NP-G384R ranged from 120 to 136%, which is higher than that of the A(H3N2)-WT, set at 100%. Specifically, the polymerase activity of the combination with PA-K615E and NP-G384R, as well as the combination with PB2-S590R and PA-K615E, was greater than that of A(H3N2)-WT. The activity levels of the combination of PB2-S590R, PA-K615E, and NP-G384R and the combination of PB2-S590R, PB1-I682V, and PA-K615E were comparable and significantly higher than A(H3N2)-WT (P < 0.05). Ultimately, the polymerase activity of the combination with PB2-S590R, PB1-I682V, PA-K615E, and NP-G384R was also significantly higher than A(H3N2)-WT (P < 0.05). These results indicated that the mouse-adapted mutations in the PB2, PB1, PA, and NP genes enhanced the polymerase activity, which may affect the replication and virulence of A(H3N2) virus in mice.

Pathogenic and replication characteristics of A(H3N2)-MA and A(H3N2)-WT strains in mice

The mouse-adapted A(H3N2) strain, A/Kansas/14/2017/H3N2-MA, was generated through three rounds of plaque purification of the viruses from lung homogenate of the MA17 passage in MDCK cells. We first investigated the pathogenic characteristics of the A(H3N2)-MA and A(H3N2)-WT strains based on the MLD₅₀ (Fig. 1A). The mice were infected with the A(H3N2)-WT strain of 10⁸ EID₅₀ and the A(H3N2)-MA strain of different infection doses (from 10³ to 10⁶ EID₅₀, to tenfold dilution), respectively. The results showed that mice infected with A(H3N2)-MA of 10³ EID₅₀ experienced a 10% decrease in body weight at 7 dpi, and later began to increase. The body weights of mice infected with 10⁴ EID₅₀ of A(H3N2)-MA decreased by approximately 20% at 8 dpi, and two of five mice died at 9 dpi (Fig. 3A and B). Most mice infected with 10⁵ EID₅₀ and 10⁶ EID₅₀ of A(H3N2)-MA lost body weights of greater than 25% at 7 dpi, and all mice succumbed at 8 and 7 dpi, respectively. In contrast, mice infected with the maximum titer of 10⁸ EID₅₀ of A(H3N2)-WT showed no significant body weight loss, and all five mice survived (Fig. 3A and B). Therefore, the MLD₅₀ value for the A(H3N2)-MA strain is 10^4.167 EID₅₀/50 μL, and for the A(H3N2)-WT is greater than 10^8.0 EID₅₀/50 μL, indicating that the virulence of A(H3N2)-MA is higher than A(H3N2)-WT in the mouse model.

Furthermore, we compared the pathogenicity of A(H3N2)-MA and A(H3N2)-WT in the mouse model infected with the same dose of 10^4.5 EID₅₀/50 μL. In the A(H3N2)-MA group, the body weights gradually decreased to approximate 77% of initial weight at 8 dpi, and two of five mice died at 8 dpi, with a survival rate of 60% (Fig. 3C and D). Conversely, mice infected with A(H3N2)-WT exhibited no body weight loss compared to the control group, and all mice survived (Fig. 3C and D). Furthermore, the mean values of lung viral RNA loads of the mice infected with A(H3N2)-MA were 10^9.82, 10^10.84, and 10^9.69 copies/mL at 3, 5, and 7 dpi, respectively (Fig. 3E), meanwhile virus loads in the nasal turbinate were 10^8.15, 10^8.91, and 10^8.17 copies/mL (Fig. 3F). In contrast, no viral RNA were detected in the lung and nasal turbinate of mice in the A(H3N2)-WT group at 3, 5, and 7 dpi. These findings indicated that the pathogenicity and replication capacity of the A(H3N2)-MA strain were significantly greater than those of the A(H3N2)-WT strain.

Tissue tropism of A(H3N2)-MA and A(H3N2)-WT in the mouse model

To evaluate tissue tropism of the A(H3N2)-MA and -WT strains, we also measured the virus RNA loads in seven tissues (including the heart, liver, spleen, kidney, intestine, stomach, and brain) from the infected mice with 10^4.5 EID₅₀ at 3, 5, and 7 dpi. The results exhibited that no viral RNA was detected in these seven tissues, regardless of the A(H3N2)-MA or A(H3N2)-WT strains (Fig. S1). These findings suggested that the tissue tropism of A(H3N2)-MA has not changed compared to the WT strain, although the replication ability of the MA strain significantly increased in the respiratory system.

Lung pathology and cytokine changes in mice infected with A(H3N2)-MA and A(H3N2)-WT

Next, we compared the pathological and inflammatory factor changes in the lungs of the mice infected with A(H3N2)-MA and A(H3N2)-WT viruses. At 7 dpi, lung tissue infected with A(H3N2)-MA showed more severe pathological damage than those with A(H3N2)-WT. In the A(H3N2)-MA group, greater infiltration of inflammatory cells into the alveoli, blood vessels, and bronchioles, as well as increased shedding of bronchiolar epithelial cells were observed (Fig. 4A). Consistently, the histopathological score for the A(H3N2)-MA group was significantly elevated compared with the A(H3N2)-WT group. But no significant difference was observed between the A(H3N2)-WT and the control groups (Fig. 4B).

We then assessed the levels of cytokines and chemokines in the mouse lungs from the A(H3N2)-MA, A(H3N2)-WT, and control groups. Pro-inflammatory factors including IL-6 (Fig. 4C), TNF (Fig. 4D), and MCP-1 (Fig. 4E) were significantly higher in the A(H3N2)-MA group than in the A(H3N2)-WT group at 3, 5, and 7 dpi. Additionally, the level of the pro-inflammatory factor, IFN-γ, was significantly higher in the A(H3N2)-MA group compared with the A(H3N2)-WT group at 5 dpi and 7 dpi (Fig. 4F). These results suggested that the infection of A(H3N2)-MA induced more severe lung pathological damage and inflammatory response than the A(H3N2)-WT.

Story Continues

By Gordon Gottsegen

Retail investors’ dip-buying remains as strong as ever despite tariffs, Middle East tensions and economic uncertainty

Tariffs, market volatility and war in the Middle East couldn’t slow down the buying spree by individual investors, as they traded a record amount of stocks in the first half of the year.

Retail investors cumulatively bought around $3.4 trillion worth of equities over the first half of 2025, according to data from Nasdaq. At the same time, they sold about $3.2 trillion worth – bringing the total traded to over $6.6 trillion.

This represents a strong bias toward buying into the market versus taking money out, despite high levels of market volatility in the first half of the year. Tariff announcements from President Donald Trump spooked markets, while investors weighed the possibility of global trade wars leading to an economic slowdown and higher inflation. The Dow Jones Industrial Average DJIA and S&P 500 SPX entered a correction, while the Nasdaq Composite COMP fell into bear-market territory. Some investors said it was “the toughest investment climate” they ever experienced.

Yet retail investors’ behavior showed that they were bullish in the face of all of this. Cumulative net retail inflows hit $137.6 billion into U.S. single stocks and exchange-traded funds the first half of the 2025, according to Nasdaq.

Data from capital-markets research firm Vanda Research differed slightly from the Nasdaq data, but it showed the same general trend. Vanda found that investors net purchased $155.3 billion worth of single stocks and ETFs in the first half of 2025. This was the largest-ever net inflow of retail investor cash since Vanda started keeping track in 2014. Inflows surpassed the previous high of $152.8 billion reached in the first half of 2021, when the meme-stock mania and pandemic stimulus checks drove hordes of everyday investors into the stock market.

In 2025, buying was driven by two things, Vanda said: the “American exceptionalism” trade and a record amount of dip-buying in response to Trump’s “liberation day” tariffs. Buzzy U.S. stocks – like Nvidia (NVDA), Tesla (TSLA) and Palantir (PLTR) – topped the charts of the most actively traded tickers throughout the first half, but retail investors also poured significant amounts of capital into index-tracking ETFs like SPDR S&P 500 ETF Trust SPY and Invesco QQQ Trust Series I QQQ.

Read more: Fourth of July holiday highlights 4 reasons ‘American exceptionalism’ isn’t going anywhere

The average daily inflow was roughly $1.3 billion, according to Vanda, which represents a 21.6% jump from the average in 2024.

This level of stock-buying hasn’t exactly hurt investors’ performance either. Vanda estimated that the average retail portfolio was up 6.2% so far in 2025, which was closely in line with the 6.1% that the S&P 500 gained in the first half of this year.

“Retail investors remain a major force in the market. Participation is at record highs, the dip-buying bias is fully intact, and engagement with single names – particularly high-beta and leveraged plays – continues to rise. Performance is holding up, and risk appetite is anything but shy. Nothing seems to stop this retail train,” Marco Iachini, Vanda Research’s senior vice president of research, wrote in a note.

-Gordon Gottsegen

This content was created by MarketWatch, which is operated by Dow Jones & Co. MarketWatch is published independently from Dow Jones Newswires and The Wall Street Journal.

  (END) Dow Jones Newswires

  07-05-25 0800ET
Copyright (c) 2025 Dow Jones & Company, Inc.

Chinese chain Luckin Coffee opened its first two U.S. locations this week, betting that mobile-only ordering and creative flavors can lure customers away from Starbucks.

Both new Luckin stores are based in Manhattan, and at the midtown location on Wednesday, Sam Liu took a sip of her jasmine cold brew.

“I’ve never tried anything like it,” she said.

Liu said she’d hoped for more seating — the small shop has only three tables — and was initially confused by Luckin’s in-app ordering system, which means customers can’t order directly from a barista.

“I thought I just order at the counter, but I realized everyone was standing around looking at their phone,” Liu said.

Luckin is China’s largest coffee chain, with more than twice as many locations as Starbucks there. Its two New York City stores are its first foray outside Asia, where it has over 24,000 locations across the region. By comparison, there are over 17,000 Starbucks in the United States.

Its CEO, Guo Jinyi, called the U.S. “a strategically important market” for the company’s expansion in a press release heralding the two new locations Wednesday. “We are excited to introduce a diverse and unique coffee experience to American consumers.”

The company, which didn’t respond to a request for comment, has touted its ambitions to expand globally but hasn’t publicly detailed its next moves in the U.S. or other markets.

The chain has gained success overseas through creative drinks like alcohol-infused coffees and fruit lattes, along with its smartphone-centric ordering model. The app-based approach makes it easier to track inventory, send personalized appeals to consumers and serve drinks quickly, said John Zolidis, an analyst who tracks Luckin and Starbucks at the brokerage firm he founded, Quo Vadis Capital.

“Luckin was able to develop an incredible muscle with regard to product innovation, and they have been very creative in China,” he said.

Zolidis said how Luckin fares on Starbucks’ home turf will depend on its ability to differentiate its menu from other major U.S. coffee chains and smaller, independent cafes. Its American lineup already includes distinctive drinks like blood orange cold brew and coconut lattes.

“These orange drinks, or one of their most successful, a coconut cloud latte — that’s how you get trial [customers] from the U.S.,” Zolidis said.

Luckin faced financial troubles during the pandemic. It was delisted from Nasdaq in 2020 after its stock plunged following an internal investigation that found an executive had falsified revenue reports. The company filed for bankruptcy in the U.S. the following year but emerged from proceedings in 2022 and its sales have soared since, reaching $4.7 billion worldwide in fiscal year 2024, a 38.4% increase from 2023.

Starbucks, by contrast, is struggling in both the U.S. and China. Its same-store sales in the U.S. declined 2% and its sales in China 8% in fiscal year 2024, and it reported in April that its quarterly profit was half of what it pulled in for the same period last year. The Seattle-based chain is reportedly looking to partially sell its business in China while revamping its U.S. strategy to focus on customer experience and human connection, in contrast with Luckin’s model.

“We veered away from, I think, owning the idea of the ‘third place,’ the coffeehouse experience, making sure that the customer was front and center,” Starbucks CEO Brian Niccol told NBC News in June.

A Starbucks spokesperson declined to comment.

Zolidis said that whereas Starbucks aims in both the U.S. and China to appeal to customers looking for higher-end coffee served in an inviting setting, Luckin has successfully positioned itself as the “everyman’s coffee” in China, with low prices and small, grab-and-go storefronts.

After taking the train in from Hoboken, New Jersey, to check out the new one in midtown, Samantha Coy said the trip was worth it. She had enjoyed Luckin in China previously and was eager to order one of its fruit drinks.

“I’m surprised Starbucks hasn’t tried to bring that over to the U.S.,” Coy said. “I hope they stay open.”

Zolidis said he thinks Luckin is well-positioned to gain a foothold in America.

“They’ve been able to operate and grow incredibly quickly in the Chinese market, much faster than I would have thought possible, and they’ve been able to sustain it and develop a strong financial model so they can fund their expansion in the U.S.,” Zolidis said. “They wouldn’t be coming here to try it if they didn’t think they had a shot of owning part of the market.”

The six-month, full-time paid program offers aspiring AI researchers from adjacent fields like physics or neuroscience a pathway into the AI industry, rather than recruiting individuals already deeply invested in AI research and work. According to Jackie Hehir, OpenAI’s research residency program manager, residents aren’t those seeking Ph.Ds in machine learning or AI, nor are they employees of other AI labs. Instead, she said in a program info session, “they’re really passionate about the space.”

So what’s in it for OpenAI? Hot talent at cut-rate prices. While the six-figure salary puts OpenAI residents in the top 5% of American workers, it’s a bargain in the rarefied world of AI, where the bidding war for talent has some companies tossing around nine-figure bonuses.

By offering a foothold into the AI world, OpenAI appears to be cultivating talent deeply embedded in the company’s mission. This strategy, spearheaded by CEO Sam Altman, has long been part of the company’s approach to retaining employees and driving innovation. One former OpenAI staffer described the employee culture to Business Insider as “obsessed with the actual mission of creating AGI,” or artificial general intelligence.

Mission driven or not, OpenAI’s residents are also compensated handsomely, earning an annualized salary of $210,000, which translates to around $105,000 for the six-month program. The company also pays residents to relocate to San Francisco. Unlike internships, the program treats participants as full-fledged employees, complete with a full suite of benefits. Nearly every resident who performs well receives a full-time offer, and, according to Hehir, every resident offered a full-time contract so far has accepted. Each year, the company welcomes around 30 residents. 

The qualifications for residents at OpenAI are somewhat unconventional. In fact, the company claims there are no formal education or work requirements. Instead, they hold an “extremely high technical bar” at parity to what they look for in full-time employees as it pertains to math and programming.

“While you don’t need to have a degree in advanced mathematics, you do need to be really comfortable with advanced math concepts,” Hehir said.

As OpenAI attempts to build talent from the ground up, its rivals, namely Meta, are pulling out all the stops to poach top AI talent with reports alleging that Meta CEO Mark Zuckerberg personally identified top OpenAI staff on what insiders dubbed “The List” and attempted to recruit them with offers exceeding $100 million in signing bonuses.

Meta’s compensation packages for AI talent can reportedly reach over $300 million across four years for elite researchers. The flood of cash has ignited what some insiders call a “summer of comp FOMO,” as AI specialists weigh whether to stay loyal to their current employers or leave for record-breaking paydays.

Zuckerberg’s methods have had some success, poaching a number of OpenAI employees for Meta’s new superintelligence team. In response to news of the employees’ departure, OpenAI’s chief research officer, Mark Chen, told staff that it felt like “someone has broken into our home and stolen something.”

Meanwhile, OpenAI CEO Sam Altman called Meta’s recruitment tactics “crazy,” warning that money alone won’t secure the best people. “What Meta is doing will, in my opinion, lead to very deep cultural problems,” Altman told employees in a leaked internal memo this week. 

Ultimately, cultivating new talent, rather than trying to outbid the likes of Meta, may prove a more sustainable path for OpenAI in its quest to stay highly mission-oriented while supporting an industry grappling with a scarcity of top-tier talent. Estimates suggest there are only about 2,000 people worldwide capable of pushing the boundaries of large language models and advanced AI research. Whether the talent cultivated by Altman and OpenAI will remain loyal to the firm remains unknown. But Altman says that AI “missionaries will beat mercenaries.”

Participants

Twenty-four healthy, community-dwelling older adults, consisting of 15 men and 9 women with a mean age of 72.1 ± 4.1 years, participated (Table 1). The eligibility criteria of this study were as follows: age ≥ 65 years; ability to walk independently without assistive devices; and absence of conditions that could significantly influence gait, such as neurological disorders (e.g., Alzheimer’s disease, Parkinson’s disease, or stroke), severe cardiovascular or respiratory impairments with symptoms during daily activities (e.g., heart failure, chronic obstructive pulmonary disease), or musculoskeletal problem that disable independent gait (e.g., joint replacement, spinal surgery, or advanced arthritis).

Sample size calculation

The required sample size was determined based on the population within-subject standard deviation (PWSD). The number of subjects was determined to estimate PWSD within 10% of the population value ((frac{1.96}{sqrt{2nleft(m-1right)}}leq0.1,;m=the;number;of;observations;per;subject)) using the variance of PSWD ((frac{{sigma }_{w}}{sqrt{2nleft(m-1right)}}, {sigma }_{w}=PWSD)) [21]. A sample size of 24 was required for all of the distances with nine or more observations per subject (m ≥ 9) except for the two longest distances (4.9 and 5-m).

Muscle mass and strength assessments

Participants underwent bioimpedance analysis using an InBody S10 device (InBody Co., Ltd., Seoul, South Korea) to determine height-adjusted appendicular skeletal muscle mass. Muscle strength was assessed by measuring handgrip and isometric knee extension strength. Handgrip strength was measured using a Takei 5401 Digital Dynamometer (Takei Scientific Instruments Co., Ltd., Niigata, Japan) in a standing position with the elbow fully extended. Isometric knee extension strength was evaluated using a TKK-5710e tension meter (Takei Scientific Instruments Co., Ltd., Niigata, Japan); during measurement, participants were seated on a chair with a dynamometer anchored to it, maintaining knee flexion at 90°. Both measurements were conducted bilaterally, with each side assessed twice and a 1-min rest period between attempts. Participants were instructed to exert maximum effort for each measurement, and the highest reading was used in the analysis. All procedures were conducted by a single trained assessor following the recommendations of Asian Working Group for Sarcopenia and the European Working Group on Sarcopenia in Older People [9, 10].

Physical performance assessments

Physical performance was evaluated using the Short Physical Performance Battery (SPPB) [22], the 30-s chair stand test [23], the five-times sit-to-stand test [24], and the timed up-and-go test [25]. All assessments were conducted by a single trained assessor in a spacious setting under consistent environmental conditions, following the protocols of the Asian Working Group for Sarcopenia and the European Working Group on Sarcopenia in Older People [9, 10].

10-m gait speed test and data acquisition

Participants walked along a 10-m walkway, which included a 2-m acceleration zone for a dynamic start and a 2-m deceleration zone at the end. They were instructed to walk at their usual pace on a hard surface while wearing comfortable footwear. The 10-m walk was repeated three times, with a minimum rest period of 2 min between trials. Recordings were captured using an Apple iPad Pro 11 2nd Generation (Apple, Inc., Cupertino, CA, USA) equipped with RGB cameras arranged perpendicularly to the walking path at a distance of 3.8 m and a height of 0.8 m. Videos were recorded in the sagittal plane (resolution: 800 × 600 pixels; 30 fps; Fig. 1).

Gait analysis using 2D pose estimation

A customized pose estimation model (ViFive, Inc., Boulder, CO, USA) was used, which tracked 14 key body points using an architecture adapted from a standard stacked hourglass model [26]. We introduced multiple objectives to enhance the context, accuracy, speed, and stability of the model, which are vital for musculoskeletal assessment. The classification model included a random forest classifier with optimized features to increase accuracy and speed while reducing the model size. Pixel-per-meter estimation used markers at 2 and 8 m (Fig. 1). The CoM of each subject was determined using the weighted sums of the body segment centers of mass (Fig. 2a).

Gait speed estimation

Gait speed was measured using two independent methods for validation: manually with a stopwatch and using pose estimation algorithms. Manually assessed speed was determined by an evaluator using a stopwatch to record the time taken for the subjects to pass by the markers set at 2 and 8 m. Pose estimation gait speed was calculated by dividing the distance covered between frames by the elapsed time using either the CoM or the leading foot as reference points. CoM-referenced measurements simulate those obtained via conventional motion capture system, whereas leading foot-referenced measurements simulate those made using walkway or pressure sensors such as GAITRite® (CIR Systems Inc., Franklin, NJ, USA).

Gait speed measurement validation

Gait speeds measured using a stopwatch and pose estimation were compared using a linear mixed-effects model, with speed over 6 (manual) or 5 m (pose estimation) per trial as the dependent variable and with subject random effect to account for multiple tests from each subject. The intraclass correlation coefficient (ICC) was used to evaluate absolute agreement between gait speed measurements obtained via the two methods for the same walking trials.

Change of uncertainty with measured distance

A 5-m walk video of a skeleton with 14 key points was extracted from each recording using our pose estimation algorithm. This was further edited by cropping at 0.1-m intervals to generate 4.9- to 1.0-m segments. One 5.0-m walk video generated two 4.9-m segments, three 4.8-m segments, and so forth, up to 41 segments for a 1.0-m walk, culminating in 861 segments of varying distance (Fig. 2b,c).

The variability of gait speed across the measured distances was defined as the within-subject standard deviation (WSD) for each measured distance, calculated as the square root of the mean-square error in a one-way analysis of variance, where groups combined subjects with distance intervals. Three gait speed data from three measurements were collected for each group to avoid underestimating within-subject variation due to overlapping distances when distance intervals are not considered. For example, for a 4.7-m walk, four gait speed measurements were obtained at 4.7-m distances (0–4.7, 0.1–4.8, 0.2–4.9, and 0.3–5.0 m), and within-subject variation at a 4.7-m distance was estimated by considering different distance intervals.

Determination of minimum required distance

To determine the minimum required distance, we utilized WSD at each measured distance. Given that confidence intervals (CI) quantify variability, we computed the half-width of the CI using WSD and the critical value corresponding to the chosen confidence level. Specifically, the 95% CI was calculated as 1.96 × WSD, and the 90% CI as 1.64 × WSD. For the measurement to be clinically meaningful, the half-width of the CI, reflecting gait speed variability, had to remain below the MCID of 0.1 m/s [27, 28]. Thus, the minimum required distance was defined as the shortest distance at which this criterion was met, ensuring that gait speed measurements remained within an acceptable range of variability.

Factors affecting gait speed variability

CoM trajectory was plotted as distance against time for each test. Linear regression analysis provided a trend line and the mean squared error (MSE) for each subject. As MSE quantifies deviations from the trend line, lower MSE values indicate less variability in gait speed, leading to a shorter minimum required distance. We investigated whether epidemiological, anthropometric, or clinical variables were associated with MSE using linear regression following Pearson’s correlation for continuous variables and point-biserial correlation for dichotomous variables to identify subject characteristics influencing the minimum required distance. All processing and statistical analyses were conducted using MATLAB R2023b (MathWorks, Natick, MA, USA) and SAS 9.4 (SAS Institute, Cary, NC, USA), with statistical significance set at p < 0.05.