About

Kyekyoon (Kevin) Kim is a professor in the Department of Bioengineering at the University of Illinois Urbana-Champaign. He holds a PhD in Applied Physics from Cornell University and has served as the Director of the Thin Film and Charged Particle Research Lab since 1980. His primary research interests include the synthesis and study of tailored organometallic, inorganic, and polymeric precursors for sol-gel processing, charged liquid cluster beam deposition, and chemical vapor deposition. He investigates methods for controlled fabrication of thin films, nanofibers, and nanoparticles using charged liquid cluster beams of precursor solutions, as well as approaches for fabricating micro- and nanoscale structures for developing new devices and material systems for applications such as controlled release, lighting, display, solar cells, photosensors, and power storage. Additionally, his work involves developing methods for creating uniform micro- and nano-spheres and capsules of biodegradable and biocompatible materials for advanced drug delivery and cell-based therapy, and developing high-power electronic devices based on III-V nitride semiconductors using plasma-assisted molecular beam epitaxy and selective area growth technology.

Research topics

• Computer Science
• Surgery
• Internal medicine
• Medicine
• Oncology
• Pathology
• Cognitive psychology
• Neuroscience
• Psychology

Selected publications

• Applying the health belief model to explore healthcare provider and patient perspectives on homologous recombination deficiency testing in cancer: A narrative review

  Cancer Treatment and Research Communications · 2026-01-01

  articleOpen access1st authorCorresponding

  BACKGROUND: Homologous recombination deficiency (HRD) testing can inform cancer treatment decisions, particularly regarding the use of poly (ADP-ribose) polymerase (PARP) inhibitors. However, uptake of HRD testing is inconsistent across healthcare settings. This review aimed to explore healthcare provider and patient perspectives on the use of HRD testing in cancer care. METHODS: A narrative review was conducted to identify studies reporting on individual views, perceptions, and experiences with HRD testing. Literature searches were performed in Embase, Medline, CINAHL, and PsycINFO for English-language articles published between 2000 and 2026. Eligible studies used quantitative, qualitative, or mixed-methods approaches. Findings were analyzed using the Health Belief Model to identify perceived benefits, barriers, and factors influencing test uptake. RESULTS: Sixteen studies were included, covering both healthcare provider and patient perspectives. The most commonly reported facilitator was the perceived benefit of HRD testing in informing treatment decisions, particularly for guiding PARP inhibitor use. Patients emphasized the importance of provider recommendations in their decision to undergo testing. Reported barriers included lack of healthcare provider knowledge, limited access to testing, inconsistent clinical guidelines, financial concerns, and low patient awareness. CONCLUSIONS: While both patients and healthcare providers recognize the value of HRD testing in cancer care, uptake is limited by informational and structural barriers. Efforts to improve access and implementation should include the development of standardized guidelines, provider and patient education, and policy strategies to ensure equitable access. Further research is needed to assess more patient perspectives and the real-world impact of HRD testing on care delivery and outcomes.

  PublisherDOI

• A Survey of On-Call Structure and Associated Compensation Mechanisms for Interventional Radiologists Practicing in the United States

  Journal of Vascular and Interventional Radiology · 2026-02-11

  article
  PublisherDOI

• Cancer epitope prediction tools and analysis pipelines in CEDAR

  Nucleic Acids Research · 2026-05-01

  articleOpen access

  Accurate identification of immunogenic cancer epitopes remains a central challenge in immuno-oncology. The Cancer Epitope Database and Analysis Resource (CEDAR, https://cedar.iedb.org/) was developed to provide comprehensive curation of experimentally validated epitopes and to foster the development of computational tools tailored to the cancer context. Recently, we released a suite of cancer-specific tools and analysis pipelines as part of the https://nextgen-tools.iedb.org/ platform, enabling users to generate, evaluate, and prioritize candidate T cell epitopes in a modular framework. Here, we present the design and functionality of these tools, describe their core methodologies, provide guidance for their use, and illustrate how they can be integrated into end-to-end pipelines. We highlight applications in cancer immunology and personalized immunotherapy by presenting practical use cases.

  PublisherDOI

• A multicenter, open-label study of RP2 oncolytic immunotherapy expressing anti–CTLA-4 in combination with second-line atezolizumab plus bevacizumab in advanced hepatocellular carcinoma (HCC) or in combination with first-line durvalumab in advanced biliary tract cancer (BTC).

  Journal of Clinical Oncology · 2026-01-10

  article

  TPS614 Background: Despite recent breakthroughs in first-line treatment of advanced HCC using immune checkpoint inhibitors (ICIs) with or without vascular endothelial growth factor inhibition, second-line options after progression on ICIs remain limited. In BTC, first-line therapy also includes ICIs, but outcomes remain suboptimal. RP2 is an enhanced potency herpes simplex virus type 1 (HSV-1) oncolytic immunotherapy that expresses GM-CSF, a fusogenic glycoprotein (GALV-GP-R − ), and an anti–CTLA-4 antibody. RP2 showed clinical activity as a single agent and when combined with ICIs in a phase 1 study in advanced solid tumors. This study evaluates the safety and efficacy of RP2 combined with atezolizumab (atezo) + bevacizumab (bev) as second-line therapy in patients (pts) with unresectable advanced HCC or RP2 combined with durvalumab (durva) in pts with unresectable advanced BTC after first-line chemotherapy is discontinued. Methods: This is an open-label, multi-cohort, phase 2 trial (NCT05733598; RP2-003). For the HCC cohort (n = 30), pts receive RP2 combined with atezo + bev. Key inclusion criteria include advanced unresectable HCC with ≥1 measurable tumor ≥1 cm (longest diameter), Child-Pugh class A, ECOG performance status of 0–1, and progression on 1 prior systemic treatment (anti–PD-1/PD-L1 as immediate prior therapy). Key exclusion criteria include untreated/incompletely treated esophageal or gastric varices with bleeding or high bleeding risk and macroscopic invasion of the tumor into any major blood vessels or main bile ducts. Pts receive intratumoral (IT) RP2 every 2 weeks (Q2W) until the 4 th dose, then Q3W thereafter. Bev is given IV at 10 mg/kg Q2W with the 1 st dose of RP2, then 15 mg/kg Q3W starting with 4 th dose and thereafter; atezo is given IV at 840 mg Q2W for 2 doses starting with the 2 nd dose of RP2, then 1200 mg Q3W thereafter. For the BTC cohort (n = 30), pts receive IT RP2 Q2W combined with durva 1500 mg Q4W after combination chemotherapy is discontinued. Pts must be on a combination of gemcitabine, platinum-containing chemotherapy, and a checkpoint inhibitor for ≥12 weeks and have documented stable disease or partial response on ≥2 scans. Pts with mismatch repair deficiency/microsatellite instability-high tumors are excluded. For both cohorts, the primary endpoint is overall response rate (ORR) per Response Evaluation Criteria in Solid Tumors version 1.1 (RECIST 1.1). Secondary endpoints include safety, ORR per RECIST 1.1 modified for use in HCC (HCC cohort only), duration of response, complete response rate, progression-free survival, and overall survival. Clinical trial information: NCT05733598 .

  PublisherDOI

• Locoregional therapy for hepatocellular carcinoma before transplant: Impact on tumor necrosis, tumor recurrence, and long-term survival.

  Journal of Clinical Oncology · 2026-01-10

  articleSenior author

  579 Background: To assess the effectiveness of different pre-transplant loco-regional therapy (LRT) approaches in patients with hepatocellular carcinoma (HCC) by tumor necrosis in liver explants, predictors of post-liver transplantation (LT) HCC recurrence and overall survival after LT. Methods: A retro analysis on consecutive patients who received LRT for HCC, within and beyond Milan criteria, prior to LT over a 12-year period. Categorical and continuous variables were examined using one-way ANOVA, while correlations among continuous variables were evaluated using correlation analysis. Linear regression was utilized to identify predictive factors. Survival outcomes were analyzed using the Kaplan-Meier method, with differences assessed via the log-rank test. Additionally, multivariate regression analysis was performed to explore the association between categorical and continuous variables and the risk of post-transplant recurrence. Results: 165 patients with HCC treated with 233 LRT sessions (median 1, range 1-5) prior to LT; 125 (76%) within Milan and 40 (24%) downstaged outside Milan, 30 (18%) had microvascular invasion in explants. Kaplan-Meier analysis showed median OS was 127 months post LT; mean 96 ± 4 months. The overall explant tumor necrosis after LT was 86%. The overall tumor necrosis with TACE+ablation, Y90, ablation, and TACE were 92, 89.3, 85.4 and 71%, respectively (p<0.05). The recurrence-free survival rates for patients treated with TACE+ablation and Y90 were both ≥90% (p=0.038) in comparison with ablation and TACE alone. Recurrence of HCC occurred in 12 patients (7.2%) over median follow-up of 11.5 months (range 4-52). Mean OS was 103 months (SD: 4, 95% CI: 95–111) for Milan In, and 92 months (SD: 9, 95% CI: 73–110) for Milan Out. Within Milan with bridging LRTs, 5 out of 125 (4%) experienced recurrence, while outside of Milan with downstaging LRTs, 7 out of 40 patients (18%) developed recurrences. Outside of Milan down-staged, pre-LRT AFP and microvascular invasion in explants were significantly and independently associated with recurrence (p=0.04, p=0.004 and p=0.02). Conclusions: The bridging and downstaging LRTs resulted in overall pathologic tumor necrosis of 86% (P>0.05). Median OS was 127 months; mean 96 ± 4 months. Mean OS was 103 months (SD: 4, 95% CI: 95–111) for Milan In, and 92 months (SD: 9, 95% CI: 73–110) for Milan Out. Post-LT recurrence-free survival in patients treated with TACE+ablation and Y90 was ≥90% (p<0.05). These findings highlight the critical impact of effective pre-transplant loco-regional therapy on long-term survival in HCC patients. Outside of Milan down-staged, pre-transplant AFP, and microvascular invasion in explants were significantly and independently correlated with post-LT HCC recurrence.

  PublisherDOI

• Abstract 5503: Cancer epitope prediction tools & analysis pipelines in CEDAR.

  Cancer Research · 2026-04-03

  article

  Abstract The identification of immunogenic cancer epitopes, including patient-specific neoepitopes and shared tumor-associated antigens (TAAs), is a central challenge for the development of effective cancer immunotherapies. To accelerate their discovery, the Cancer Epitope Database and Analysis Resource (CEDAR), a comprehensive resource for immuno-oncology, curates epitope data from the literature and develops tailored computational tools. Building on this foundation, we introduce the modular Next-Generation IEDB Tools (NGT) platform (nextgen-tools.iedb.org/) that integrates a wide array of cancer-focused computational tools. The NGT platform allows users to construct, save, and share customized, reproducible, end-to-end computational pipelines for tumor antigen discovery. This architecture enables systematic prioritization of epitope candidates by applying multiple, sequential filtering criteria based on predicted and calculated relevant immune features, such as antigen expression, antigen presentation, self-similarity, and immunogenicity. Key tools include the Mutated Peptide Generator (MPG), which translates genomic variants (e.g., SNVs, indels) into candidate neoepitope sequences; Peptide Expression Annotation (PepX), which integrates public RNA-Seq data from resources like TCGA and GTEx to quantify antigen-encoding transcript abundance in tumor tissue; the Patient-Specific Presentation Metric (PHBR), a metric that estimates the likelihood of a mutation being presented by a patient's specific MHC Class I alleles; ICERFIRE (via Peptide Variant Comparison, PVC), a robust immunogenicity model that predicts the T-cell recognition potential of neoepitopes; PEPMatch, a tool to filter out candidate neoepitopes that are highly similar to self-peptides, which can indicate potential tolerance or autoimmunity risks; and Cluster, which groups highly similar peptide sequences to reduce redundancy and focus on the most representative candidates for experimental validation. The CEDAR computational tools and integrated pipeline architecture on the NGT platform provide cancer immunologists with a flexible, user-friendly, and state-of-the-art resource. This comprehensive framework accelerates the translation of genomic and transcriptomic sequencing data into clinically actionable epitope candidates. Here, we present how these integrated tools can be applied to patient-level analyses, enabling personalized identification and prioritization of tumor epitopes to guide cancer vaccine design and immunotherapy development. Citation Format: Ibel Carri, Jason Greenbaum, Zhen Yan, Kevin Kim, Haeuk Kim, Ashmitaa Logandha Premlal, Daniel Marrama, Nina Blazeska, Hannah K. Carter, Morten Nielsen, Alessandro Sette, Bjoern Peters, Zeynep Kosaloglu-Yalcin. Cancer epitope prediction tools & analysis pipelines in CEDAR [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 1 (Regular Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(7 Suppl):Abstract nr 5503.

  PublisherDOI

• Synthesis and infrared nanospectroscopy of MXene materials, application to the detection of siloxane compounds

  theses.fr (ABES) · 2025-12-18

  articleOpen access1st authorCorresponding

  Cette thèse étudie les propriétés optiques infrarouges des MXenes par le biais d'une caractérisation à plusieurs échelles, établit des relations structure-propriété fondamentales et présente des applications pratiques. Le chapitre 1 décrit les matériaux bidimensionnels, positionnant les MXènes comme de nouveaux carbures et nitrures de métaux de transition 2D avec une conductivité métallique élevée et une chimie de surface accordable. Le chapitre 2 développe les bases théoriques de la caractérisation optique multi-échelle et introduit des techniques avancées de microscopie en champ proche comme la microscopie optique en champ proche à balayage de type diffusion (s-SNOM) et la microscopie à force photo-induite (PiFM) pour l'investigation à l'échelle nanométrique au-delà des limites de la diffraction. Le chapitre 3 démontre le contrôle systématique des propriétés infrarouges par l'ingénierie de la composition, révélant que le rapport carbone/azote dans les MXènes Ti3C(2-y)NyTx permet une modulation de la réflectance et de l'émissivité. Les compositions riches en carbone présentent des capacités exceptionnelles de blindage IR, tandis que l'incorporation d'azote augmente pour les applications de gestion thermique. De plus, l’étude infrarouge in-situ des intercalations ioniques dans le MXene Ti3C2Tx montre des réponses optiques dynamiques pendant les processus électrochimiques, fournissant des informations sur les modifications structurales en temps réel et les mécanismes de transfert de charge. La PiFM est utilisée dans le chapitre 4 pour la caractérisation de la chimie des surfaces à l'échelle nanométrique, mettant en évidence des terminaisons de surface natives et une contamination par le siloxane non détectée par méthode conventionnelle. L'étude est étendue à la détection de l'oxydation du Nb2CTx et à la liaison préférentielle du siloxane aux plans basaux composés de terminaison hydroxyle. Dans le Chapitre 5, nous appliquons les MXenes aux plateformes de détection par Résonance Plasmonique de Surface. Le MXene Ti3C2Tx améliore le couplage SPR et permet la détection du (3-aminopropyl)triéthoxysilane (APTES). L'approche de caractérisation optique multi-échelle développée met en lumière les performances et la versatilité de cette famille émergente de matériaux 2D.

  Publisher

• Collagen fibril formation at the plasma membrane occurs independently from collagen secretion

  Wellcome Open Research · 2025-08-29

  preprintOpen access

  Background: Collagen fibrils are the primary supporting scaffolds of vertebrate tissues, but the mechanism of assembly is unclear. Methods: Here, using CRISPR-tagging of type I collagen, high-resolution light imaging, and SILAC labelling, we elucidated the cellular mechanism underlying the spatiotemporal assembly of collagen fibrils in cultured fibroblasts. Results: Our findings reveal the multifaceted trafficking of collagen, including constitutive secretion, intracellular pooling, and plasma membrane-directed fibrillogenesis. Notably, we differentiated the processes of collagen secretion and fibril assembly and identified the crucial involvement of endocytosis in the regulation of fibril formation. By employing Col1a1 knockout fibroblasts, we demonstrated the incorporation of exogenous collagen into the nucleation sites at the plasma membrane through these recycling mechanisms. Conclusions: Our study sheds light on a complex and previously unidentified collagen assembly process and its regulation of health and disease. Mass spectrometry data were available via ProteomeXchange with the identifier PXD036794.

  PublisherDOI

• Lysosomal permeabilization by Group A <i>Streptococcus</i> releases proteins into the macrophage cytosol

  bioRxiv (Cold Spring Harbor Laboratory) · 2025-10-25

  preprintOpen access

  ABSTRACT The human-specific bacterial pathogen Group A Streptococcus (GAS) is a significant cause of morbidity and mortality due to its ability to cause severe invasive infection. Although macrophages are important for controlling GAS infection, we and others have demonstrated that GAS can persist in macrophages by perforating the phagolysosome using the pore-forming toxin streptolysin O (SLO). In this study, we examined how phagosomal perforation releases lysosomal and bacterial proteins into the cytosol and alters cytosolic protein content in the macrophage. Using IL-1β as a measure of intracellular pathogen detection, we confirmed that cytosolic preparations from macrophages infected with either wild-type (WT) or SLO-deficient (ΔSLO) bacteria contained new proteins that are absent in uninfected cytosol controls. Proteomic analysis revealed distinct cytosolic protein profiles in both WT- and ΔSLO-infected macrophages. M1 protein was detected only in the cytosol of WT-infected macrophages and corresponded with the IL-1β response, indicating SLO-mediated release of M1 protein from the phagosome, and providing a mechanism for cytosolic recognition of this virulence factor. Unexpectedly, cytosolic extracts of both WT- and ΔSLO-infected macrophages contained histone proteins H1–H4, suggesting that nucleosomal complexes are released into the cytosol during GAS infection. Our work reveals both a mechanism for the activation of the inflammatory response on a cellular level, and the surprising consequences of phagosomal perforation in GAS infections. These responses may collectively contribute to the pathologies observed during severe invasive GAS infection, and can help inform therapies aimed at improving macrophage function and patient outcomes.

  PublisherDOI

• A multicenter, open-label study investigating RP2 oncolytic immunotherapy in combination with second-line systemic atezolizumab plus bevacizumab in patients with locally advanced unresectable or metastatic hepatocellular carcinoma (HCC).

  Journal of Clinical Oncology · 2025-01-27

  article1st authorCorresponding

  TPS649 Background: Despite advances in unresectable HCC treatment, long-term survival rates remain poor. The combination of atezolizumab (Atezo) plus bevacizumab (Bev) is approved as frontline therapy for advanced HCC, but only a minority of patients (pts) respond, and secondary resistance usually occurs within months. HCC has an immune-suppressed tumor microenvironment (TME) mediated by the expression of immune checkpoint signals and angiogenesis pathways, which may contribute to therapeutic resistance. RP2 is an enhanced potency oncolytic herpes simplex virus type 1 (HSV-1) that expresses GM-CSF, a fusogenic glycoprotein (GALV-GP-R–), and an anti–CTLA-4 antibody-like molecule. RP2 showed preliminary clinical activity alone or combined with anti–PD-1 in a phase 1 study in pts with advanced solid tumors. The direct oncolytic effect coupled with immune stimulation by RP2 in the TME is intended to provide systemic antitumor activity and synergize with anti–PD-1/PD-L1 agents, such as Atezo. Preclinical data have demonstrated improved distribution of oncolytic HSV within tumors when administered with Bev, supporting the clinical combination of RP2 with Bev. This study will evaluate the safety and efficacy of RP2 combined with Atezo plus Bev as second-line systemic therapy for unresectable advanced HCC (NCT05733598). Methods: This is an open-label, single arm, phase 2 trial. Up to 30 pts will be enrolled and receive RP2 in combination with Atezo plus Bev. Key inclusion criteria include advanced unresectable HCC with ≥1 measurable tumor of ≥1 cm in longest diameter, Child-Pugh class A, an Eastern Cooperative Oncology Group performance status of 0 to 1, and progression on 1 prior systemic treatment, which must have included a PD-1/PD-L1–directed agent. Key exclusion criteria include untreated/incompletely treated esophageal and/or gastric varices with bleeding or at high risk for bleeding and macroscopic invasion of the tumor into any major blood vessel(s) and/or main bile ducts. Pts will receive intratumoral RP2 Q2W for 4 doses, then Q3W for up to 4 doses. Bev will be given at 10 mg/kg Q2W starting with the first dose of RP2, then at 15 mg/kg Q3W starting with cycle 4; Atezo will be given at 840 mg Q2W for cycles 2 and 3, then at 1200 mg Q3W starting with cycle 4. Pts will receive treatment until confirmed progressive disease, loss of clinical benefit, or unacceptable toxicity. The primary endpoint is overall response rate (ORR), defined as the proportion of pts achieving a best overall response of complete response or partial response per Response Evaluation Criteria in Solid Tumors (RECIST) version 1.1 as modified for this study. Secondary endpoints include safety, ORR using HCC-modified RECIST, duration of response, complete response rate, and progression-free survival. Clinical trial information: NCT05733598 .

  PublisherDOI

Recent grants

• The effect of p66 and diet on stem cell regeneration and stress resistance

  NIH · $24.0M · 2007–2018

Frequent coauthors

• Russell K. Pachynski

  37 shared

• Muhammad A. Saeed

  Washington University in St. Louis

  36 shared

• Daniel L.J. Thorek

  Washington University in St. Louis

  31 shared

• Nadeem A. Sheikh

  30 shared

• Ariel Borkowski

  30 shared

• Todd A. Fehniger

  30 shared

• Kavita Rawat

  Dartmouth College

  29 shared

• Brian A. Van Tine

  Washington University in St. Louis

  29 shared

Labs

• Kevin Kim Research GroupPI

Education

• Ph.D., Bioengineering

  University of Illinois Urbana-Champaign

  2005

• M.S., Bioengineering

  University of California, San Diego

  2000

• B.S., Bioengineering

  University of California, San Diego

  1999

Awards & honors

• ECE Faculty Outstanding Teaching Award (2004)
• Nominated for the ECE Faculty Outstanding Teaching Award (19…
• Research Honors Fellow, American Institute for Medical and B…
• Arnold O. Beckman Research Award, University of Illinois (De…
• Outstanding Poster Presentation Award, Controlled Release So…