The project originated within the Rutgers Cancer Institute’s blood cancer clinic, where patients routinely faced waits stretching up to three hours between initial check-in and starting their infusion treatments. Recognizing the emotional and physical toll this imposed on patients undergoing strenuous cancer regimens, the research team sought a data-driven method to identify and alleviate such bottlenecks. By fusing operational research principles with detailed patient movement logs and time-stamped electronic health records, they constructed a digital twin of the clinic. This three-dimensional, animated simulation accurately recreated the clinic environment and patient journey, enabling exhaustive experimentation in a risk-free virtual arena.

What distinguished this digital twin model from traditional process analyses was its statistical validation against months of real patient data that the simulation itself had never encountered before. This rigorous validation established a high degree of confidence that simulated interventions could reliably forecast real-world impacts. Through this, the research dispensed with initial assumptions—such as adding more nursing staff—to pinpoint fundamental constraints in clinic throughput. For example, it became clear that increasing the number of nurses only marginally reduced visit times, shaving under a minute off average patient stay.

Instead, the study illuminated that the primary choke points were external to staff availability. Delays originated mainly from the lab processing of blood samples—a critical prerequisite to initiating infusions—which was conducted off-site and took approximately ninety minutes. Additionally, the clinic’s unified queue system led to inefficient patient sequencing, such as shorter-duration blood checks being delayed behind patients receiving eight-hour infusions. The simulation underscored that addressing these inefficiencies could lead to dramatic reductions in patient visit times—on the order of 75 to 90 minutes—even amid a 20% increase in patient volume.

In response to these insights, the clinic implemented significant operational changes. Lab processes were relocated on-site, accelerating blood work turnaround from roughly an hour and a half to under thirty minutes. Furthermore, they instituted a “fast track” separate from traditional cancer treatments, designed to streamline supportive care procedures such as transfusions and quick blood tests. Notably, this fast track feature had been available but underutilized in existing scheduling software prior to the study. The resultant workflow enhancements nearly doubled daily infusion patient throughput, scaling from around 50 to approximately 80 without compromising quality or safety.

The success story at Rutgers underscores the transformative power of computational modeling in healthcare settings, particularly those characterized by complex, multi-step patient pathways constrained by limited resources. The research highlights that while conventional wisdom might prompt institutions to augment human resources, a more fruitful approach lies in reengineering systemic bottlenecks through data-driven simulation. Each clinic’s unique layout, staffing, and patient population means that solutions must be tailored rather than transplanted wholesale, but the framework of virtual, simulation-based optimization offers a replicable blueprint.

Andrew Evens, deputy director for clinical services at Rutgers Cancer Institute and senior author of the study, emphasized the emotional and physical complexities cancer patients endure and the corresponding imperative to make their clinical experiences as efficient and compassionate as possible. Evens’ dual expertise—as a physician and holder of an executive MBA—fueled the partnership with Rutgers Business School. This interdisciplinary collaboration integrated clinical knowledge with advanced supply chain management techniques, leading to the project’s success.

Graduate students played a pivotal role, embedded in the clinic to meticulously observe and document patient movements and timing at each stage. These granular data points were combined with electronic health records in probabilistic models to generate detailed patient flow distributions. By analyzing patterns at every juncture—arrival, check-in, lab work, infusion, and check-out—the team constructed a highly granular and dynamic simulation reflective of real operations rather than theoretical approximations.

The validated digital twin empowered planners to test various hypothetical adjustments—such as rescheduling appointments, reallocating staff, or rerouting patient flows—without disrupting actual clinical operations. For instance, simulations revealed that equalizing appointment loads throughout the day rather than frontloading or clustering them reduced peak congestion, thereby smoothing patient experiences. These virtual trial runs eliminated guesswork and enabled strategic, evidence-based decision-making in optimizing clinic function.

With the transition of Rutgers Cancer Institute into the new Jack & Sheryl Morris Cancer Center—a state-of-the-art facility with distinct floors dedicated to blood draws, doctor visits, and infusion treatments—the previously solved workflow puzzles have evolved, presenting fresh operational challenges. Evens anticipates reengaging the business school team to create updated simulation models for this redesigned environment to maintain and improve efficiency gains.

The broader implications of this research resonate across the healthcare landscape, where patient throughput, wait times, and resource allocation remain persistent challenges. By harnessing computational simulations as virtual laboratories for operational experimentation, medical centers can make transformative improvements that enhance patient experience, increase treatment capacity, and potentially reduce costs. While the Rutgers model is site-specific, the underlying principle—that complex healthcare processes can be optimized through data-intensive, simulation-driven analysis—has universal application.

As healthcare systems worldwide grapple with increasing patient volumes and strained resources, this research charts a promising path. Digitally replicating clinical environments and iteratively testing process optimizations prior to implementation offers a proactive, precise, and patient-centric strategy. Rutgers’ success demonstrates that technological innovation combined with interdisciplinary collaboration can unravel even the most entrenched systemic inefficiencies in medical care delivery.

Subject of Research: Not available

Article Title: Enhancing efficiency and workflow in oncology outpatient services through simulation-based optimization

News Publication Date: 26-May-2026

Web References:
– https://link.springer.com/article/10.1007/s10479-026-07181-2
– http://dx.doi.org/10.1007/s10479-026-07181-2
– https://cinj.org/
– https://www.business.rutgers.edu/
– https://www.rwjbh.org/treatment-care/cancer/our-cancer-centers/jack-sheryl-morris-cancer-center/

References: Annals of Operations Research (2026). “Enhancing efficiency and workflow in oncology outpatient services through simulation-based optimization,” DOI: 10.1007/s10479-026-07181-2.

Keywords: Oncology outpatient services, computational simulation, digital twin, workflow optimization, cancer treatment efficiency, blood cancer clinic, operational research, patient flow management, healthcare analytics, infusion therapy throughput, laboratory turnaround time, healthcare process improvement.

The central focus of this research is PHGDH, a metabolic enzyme that plays a pivotal role in cancer cell biology. PHGDH catalyzes the first step in serine biosynthesis, a non-essential amino acid vital for rapid tumor growth. Many cancers, including roughly half of colorectal cancers and 40% of breast cancers, exhibit abnormally high levels of PHGDH, making it an attractive target for drug development. However, until now, the potential of PHGDH in cancer immunology remained unexplored.

Rutgers researchers, led by Professor Zhaohui Feng and assistant professor Juan Liu, unveiled a surprising noncanonical function of PHGDH that extends beyond its metabolic role. They discovered that PHGDH directly stimulates the production of PD-L1, a protein on tumor cells that inhibits immune attack by binding to PD-1 receptors on T-cells. PD-L1 blockade underpins the mechanism of several FDA-approved immunotherapy drugs, including checkpoint inhibitors, which aim to unleash the immune response against tumors.

Intriguingly, the team demonstrated that PHGDH’s promotion of PD-L1 expression operates independently of its enzymatic activity related to serine synthesis. By engineering mutated forms of PHGDH incapable of metabolic function, they confirmed that these inactive variants still elevated PD-L1 levels, revealing a dual role for PHGDH in tumor growth and immune evasion. This finding challenges the conventional understanding of metabolic enzymes as single-function entities.

Current experimental PHGDH inhibitors are designed to shut down the enzyme’s metabolic function, thereby starving tumors of serine. However, these drugs do not eliminate the PHGDH protein itself, leaving its immune-modulating activity intact. This distinction is critical because it means that such inhibitors may not fully counteract cancer’s ability to hide from the immune system despite reducing tumor fuel supply.

To address this complexity, Feng and his team hypothesized that combining PHGDH metabolic inhibitors with immunotherapy drugs targeting PD-L1 or PD-1 could have a powerful, synergistic effect. Testing this dual approach in mouse models, they observed a near threefold increase in survival compared to either treatment alone. Approximately 50% of mice receiving both drugs survived 60 days without signs of toxicity, whereas survival in single-treatment groups hovered around 20%, and untreated controls saw no long-term survivors.

These preclinical results suggest that attacking tumors on two fronts—metabolic deprivation and immune unmasking—can overcome resistance mechanisms that limit current immunotherapy efficacy. If translated into clinical practice, this approach holds promise for dramatically improving outcomes in cancers characterized by high PHGDH expression.

Beyond therapeutic innovation, the study identified PHGDH as a potent predictive biomarker for immunotherapy responsiveness. Analyzing existing clinical data, the researchers found that cancer patients with tumors expressing elevated PHGDH levels were significantly more sensitive to anti-PD-1 therapies. This biomarker could prove instrumental in personalizing treatment decisions, sparing patients unlikely to benefit from unnecessary side effects and costs associated with immunotherapy.

Feng emphasized this translational potential, explaining that quantifying PHGDH levels in tumors might soon guide oncologists in selecting patients most likely to respond to checkpoint inhibitors. Such biomarker-driven approaches advance the precision oncology paradigm, optimizing therapeutic efficacy on an individual basis.

The discovery integrates metabolic biology and immuno-oncology in a novel conceptual framework, highlighting how enzymes traditionally classified by their metabolic function can have multifaceted roles in cancer pathogenesis. It challenges researchers to re-evaluate the complexity of tumor biology and the interconnectedness of metabolic and immune pathways.

As PHGDH inhibitors proceed through preclinical development, regulatory approval will be essential before this combination strategy can be tested in human clinical trials. Meanwhile, ongoing biomarker validation studies aim to confirm the robustness of PHGDH expression as a predictive tool across larger and more diverse patient cohorts.

Funding from the National Institutes of Health, the New Jersey Commission on Cancer Research, the New Jersey Health Foundation, and Ludwig Research Support was critical to advancing this multidisciplinary investigation. This study exemplifies the collaborative synergy needed to translate bench discoveries into potential life-saving therapies.

This groundbreaking research opens new horizons not only for colorectal and breast cancers but also potentially for other malignancies where PHGDH plays a role. It underscores the urgency of exploring enzyme functions beyond metabolism and integrating these insights into innovative treatment paradigms that enhance immunotherapy’s reach.

As the oncology community awaits further clinical validation, patients and physicians alike can find hope in these advances. Understanding the dual roles of PHGDH may ultimately transform how cancers are treated, moving us closer to personalized, highly effective therapies that address both tumor growth and immune evasion.

Subject of Research: Animals
Article Title: Targeting the noncanonical function of metabolic enzyme PHGDH in driving PD-L1 expression and cancer immune evasion
News Publication Date: 28-Mar-2026
Web References: http://dx.doi.org/10.1016/j.xcrm.2026.102704
Keywords: Cancer immunotherapy, PHGDH, PD-L1, checkpoint inhibitors, metabolic enzyme, serine biosynthesis, immune evasion, biomarker, colorectal cancer, breast cancer, combination therapy

Researchers at Rutgers Cancer Institute, an NCI-designated Comprehensive Cancer Center, have pioneered an innovative computational methodology that promises to settle this debate decisively. Their newly developed tool, PRISM (Precise Identification of Species of the Microbiome), is a breakthrough in distinguishing authentic microbial signals embedded in human tumor sequencing data from those arising as artifacts or contamination. Publication of their detailed findings in the journal Cancer Cell marks a milestone in cancer microbiome research.

The principal challenge PRISM addresses is deceptively simple yet scientifically complex: differentiating true microbial DNA sequences within tumor samples from extraneous microbial contamination ubiquitous in lab environments. Given that microbes inhabit every conceivable surface including skin, breath, laboratory reagents, and even airborne particulates, contamination is an omnipresent threat frustrating attempts to accurately characterize tumor-associated microbiomes. A concrete illustration of this problem is the detection of microbial fragments that may have nothing to do with the tumor itself but instead infiltrated samples during routine laboratory handling.

PRISM’s architecture incorporates a multi-tiered approach that first performs rapid preliminary screens to catalog potential microbial sequences from raw sequencing data primarily intended for human genetic analysis. This step is followed by rigorous filtering to eliminate residual human sequences masquerading as microbial. Subsequently, PRISM undertakes complete sequence alignments using comprehensive microbial reference databases to accurately characterize candidate microbes. The crowning element is a machine-learning algorithm meticulously trained on an extensive dataset of 833 samples across more than 200 studies with validated microbial compositions, allowing PRISM to predict with over 90% sensitivity and specificity which microbial sequences reflect true presence versus contamination.

One of the great advantages of PRISM lies in its ability to extract meaningful microbial insights retrospectively from massive repositories of existing human genomic and transcriptomic datasets. Conventional microbiome sequencing is costly, requiring specific sample collection protocols and extensive wet-lab experimentation. PRISM cleverly repurposes standard tumor sequencing data, unlocking a treasure trove of latent microbial information without additional expense or specialized sample requirements. This paradigm shift democratizes tumor microbiome investigations by leveraging completed human sequencing efforts, propelling research forward at unprecedented scale and speed.

A comprehensive meta-analysis utilizing PRISM on nearly 4,400 tumor samples from 25 cancer types—sourced from The Cancer Genome Atlas and the Clinical Proteomic Tumor Analysis Consortium—yielded fascinating insights that realigned tumor microbiome profiles with biological expectations. Consistently, cancers arising from microbe-rich tissues such as the head and neck region, gastrointestinal tract, and cervix exhibited stronger microbial signals. In stark contrast, internal tumors from organs typically shielded from environmental microbes presented minimal microbial DNA, challenging prior reports that suggested widespread tumor-resident microbiomes. This observation reinstates fundamental microbial biology principles regarding tissue-specific colonization.

PRISM additionally illuminated the pervasive influence of laboratory contaminants in previous tumor microbiome studies. Many microbes reportedly abundant in tumors outside classical microbe-dense sites were frequently identified as common lab contaminants, thus demystifying misleading conclusions attributing robust microbiomes to tumors anatomically sequestered from the external environment. This finding underscores the critical necessity of stringent contamination controls and computational deconvolution for credible microbial detection in molecular oncology.

An illuminating case study from the research focused on pancreatic cancer samples. PRISM stratified a subset of these tumors as harboring true microbial inhabitants, notably Escherichia coli strains capable of producing colibactin, a genotoxin associated with DNA damage. This microbial presence correlated with distinctive molecular changes involving glycoprotein modifications within the tumor microenvironment. Specifically, these glycosylation shifts affected pathways involved in fibrosis—a hallmark of pancreatic cancer characterized by dense, fibrotic stroma that impedes drug delivery and immune infiltration. Such mechanistic linkages hint at microbial contributions to tumor pathophysiology, though causality remains to be fully established.

Furthermore, correlational analyses revealed that patients with histories of heavier smoking exhibited higher microbial abundances in their tumors, suggesting lifestyle factors may modulate tumor microbiomes and consequently influence disease trajectory and therapeutic responses. This intersection of environmental exposures, microbial ecology, and tumor biology represents a fertile ground for future research unlocking novel biomarkers and therapeutic targets.

While PRISM cannot singlehandedly prove whether detected microbes are oncogenic drivers or passive passengers, it sharpens the focus on biologically plausible host-microbe interactions by filtering out spurious signals. By enabling high-confidence detection of microbial taxa within tumors using only human sequencing data, the tool empowers researchers to formulate targeted hypotheses and design downstream validation experiments. This refined analytical precision significantly advances the quest to personalize microbiome-informed cancer treatment strategies.

The broader implications of PRISM extend beyond oncology. Given the tool’s adaptability to any genomic sequencing dataset, it holds promise for unraveling microbiome roles across a spectrum of diseases where microbial influence is suspected—gastrointestinal disorders, autoimmune diseases, and beyond. Its open-access availability to the academic community via GitHub accelerates collaborative innovation, although Rutgers has sought intellectual property protection for commercial applications.

In sum, PRISM represents a transformative convergence of computational biology, genomics, and microbiology. By merging machine learning with meticulous sequence alignment workflows, it transcends prior limitations and delivers unprecedented clarity on microbial presence within tumors. As this technology disseminates through the research ecosystem, it holds potential to reshape our molecular understanding of cancer and harness the microbiome’s therapeutic potential with renewed rigor.

The development of PRISM marks a pivotal advance in the rigorous detection and interpretation of microbial signatures in cancer genomics. Its capacity to disentangle true microbial residents from contamination artifacts not only clarifies longstanding controversies in tumor microbiome research but also provides a scalable tool to unlock mechanistic insights. This breakthrough empowers scientists to chart hitherto obscured host-microbe interactions across cancer types, paving the way toward microbiome-informed diagnostics and precision oncology therapies that could ultimately improve patient outcomes.

Subject of Research: Not applicable

Article Title: Reliable detection of Host-Microbe Signatures in cancer using PRISM

News Publication Date: 5-Feb-2026

Web References:

• Cancer Cell article
• DOI link

References: Rutgers Cancer Institute study published in Cancer Cell, 2026

Keywords: Cancer, Microorganisms