How 3D Models Are Strengthening Drug Repurposing in Ovarian Cancer
At the University of Technology Sydney, Amy Sarker and the Marsh lab are using RASTRUM-enabled 3D models to strengthen drug repurposing pipelines in ovarian cancer, building on earlier translational success and advancing new studies in biologically relevant preclinical models.
By the time a new ovarian cancer therapy reaches clinical trial, years of high-stakes decisions have already shaped its chances.
It starts in the lab, with difficult questions about which drug candidates are worth pursuing, which models are strong enough to de-risk those choices, and how to generate evidence that is both biologically relevant and practical enough to move a program forward.
At the University of Technology Sydney, that challenge is central to the work of Amy Sarker, PhD.
As an early-career researcher in the lab of Professor Deborah Marsh, Amy is helping advance drug repurposing workflows for ovarian cancer that rely on New Approach Methodologies, or NAMs, including in silico screening, 3D bioprinted cancer models, and patient-derived systems. The goal is not simply to find interesting signals. It is to identify candidates with the strongest potential to translate.
That work is part of a broader translational effort that has already produced a meaningful clinical milestone. In research published in 2024 led by Professor Nikola Bowden at the University of Newcastle, and collaboratively with Prof Caroline Ford’s lab at UNSW, the Marsh lab worked to identify efavirenz as a potential treatment for high-grade serous ovarian cancer using an in silico screening platform, 2D cell lines, RASTRUM 3D bioprinted models, and patient-derived organoids. That study included Christine Yee, now at Inventia Life Science, among its authors. The paper concluded that efavirenz may be a viable treatment option for HGSOC, and efavirenz is now in an open Phase II clinical trial in platinum-resistant or heavily pretreated high-grade serous ovarian cancer.
Amy Sarker’s research is now building on that earlier success.
Building a More Translational Drug Repurposing Pipeline
“I am a huge fan of NAMs in cancer research,” Amy says. “From honours through to my PhD to my postdoctoral research, my work has centred on tumour microenvironment, mechanobiology and developing 3D cell models, so I’ve always approached cancer biology through this lens.”
That perspective is especially relevant in ovarian cancer, where treatment resistance remains a major challenge and targeted therapies are still limited across multiple subtypes.
“A significant priority of our research is expediting the translation of promising preclinical findings into actionable outcomes for ovarian cancer,” Amy explains. “It is a well-documented problem in cancer research that many drugs do not progress into the clinic.”
Part of that problem is biological mismatch. Traditional 2D cultures and animal models have contributed enormously to the field, but they do not always capture the complexity of human disease.
“Cells can respond very differently to treatments depending on whether they're in a 2D or 3D microenvironment,” Amy says. “Given that human tissues are not flat plastic dishes, we need models that reflect the physical and biological reality of how cells exist and interact in vivo.”
That conviction runs through the Marsh lab’s broader approach. Deborah sees particular opportunity for NAMs in the drug repurposing space, where human safety data already exists and the challenge is to build enough confidence in efficacy to justify moving a candidate into the clinic.
“Especially in the drug repurposing space where human safety and efficacy testing has already been undertaken, NAMs present particular opportunities when laying the groundwork that will enable decisions to be made about whether to take a repurposed drug to a clinical trial,” Deborah says.
Advancing an Established RASTRUM Workflow
In the Marsh lab, RASTRUM is part of a staged translational workflow designed to bridge computational discovery and testing in more complex patient-derived systems.
That broader pipeline was already in place before Amy joined the group. The 2024 efavirenz study used RASTRUM-generated 3D hydrogel-encapsulated ovarian cancer models alongside patient-derived organoids and helped establish a strong preclinical workflow for drug repurposing in ovarian cancer.
Amy’s role has been to help extend and strengthen that foundation through newer studies. She is an author on the Marsh lab’s 2026 ibrutinib paper in ovarian clear cell carcinoma, which validated drug response in both 2D and 3D bioprinted models using the RASTRUM platform. She is also an author on the 2026 ivacaftor paper led by Prof Caroline Ford’s laboratory in high-grade serous ovarian cancer, which tested the drug in 2D, 3D bioprinted formats, and patient-derived organoids.
“I think that the RASTRUM is a great complement to a drug repurposing pipeline, offering a reproducible and biologically relevant validation step,” Amy says. “Our in silico computational methods can generate large lists of candidates, which are rigorously triaged for further in vitro screening. The RASTRUM 3D models can provide added confidence that the drug candidates we select could be successfully translated into the clinic.”
That role is especially important in ovarian cancer, where researchers need a practical way to prioritize the strongest candidates before moving into patient-derived material.
“Patient material is always precious and limited,” Deborah says. “In many research labs, it is not always feasible or possible to undertake large-scale drug screening in organoids or tumoroids for this reason. RASTRUM-enabled 3D models provide that important intermediate step between in silico discoveries and confirmation in primary patient material. In this way, the drugs most likely to succeed can be prioritised for testing on primary human tissue.”
Rather than asking a single model to answer every question, the group has built a pipeline in which each system adds confidence at the right point.
Reproducibility, Throughput, and a More Standardized Workflow
For Amy, one of RASTRUM’s biggest advantages has been the ability to test drug candidates across multiple ovarian cancer models in a consistent, scalable format.
“One of the real strengths of the RASTRUM is its reproducibility and throughput,” she says. “Being able to systematically print and test drug candidates across multiple cell lines in a consistent, reliable format has been far more achievable than if we relied on conventional approaches alone.”
That matters in drug repurposing, where comparing responses across ovarian cancer subtypes and genetic backgrounds depends on minimizing experimental variability. It also matters operationally.
“There is a significant amount of time and effort that goes into manually establishing 3D cultures, so the RASTRUM frees up a lot of physical and mental capacity for us,” Amy adds.
The team has also built a standardized 3D drug assay workflow around the platform. For each cell line, they first optimize the treatment window based on aggregate formation and viability over time, ensuring that cells are in exponential growth when drug is applied. “This has been quite straightforward to implement with the RASTRUM Cloud interface,” Amy says.
That combination of throughput, reproducibility, and workflow control has made it easier for the lab to evaluate candidates systematically and generate data they can compare with greater confidence.
Seeing What 2D Can Miss
One of the strongest arguments for 3D models is not simply that they better resemble tissue in theory. It is that they can reveal biologically meaningful responses that simpler systems fail to capture.
Amy has seen this firsthand in the lab’s repurposing work.
“There was an instance while validating the efficacy of another repurposed drug,” she says. “Preliminary testing with standard 2D drug assays indicated that our drug candidate was not very effective in reducing the viability of high-grade serous ovarian cancer cells. However, RASTRUM-bioprinted models showed that the drug dismantled 3D cellular aggregates, even at lower doses.”
That result changed how the team thought about the candidate.
“While the drug was not overtly cytotoxic, 3D cultures revealed a possible alternative mechanism, and this finding prompted us to re-assess the use of this drug as a potential anti-metastatic agent that could be used in conjunction with standard-of-care therapies to contain aggressively spreading cancers,” Amy explains.
This is one of the clearest advantages of more advanced models in drug repurposing. They do not just strengthen confidence in expected biology. They can also uncover response patterns and mechanisms that might be missed entirely in 2D.
Amy’s Contribution to the Next Phase of the Work
Amy’s work now sits within that broader trajectory, helping extend the Marsh lab’s RASTRUM-enabled drug repurposing pipeline into new ovarian cancer questions and preclinical models.
In the 2026 ibrutinib study, the group identified ibrutinib as a potential treatment for ovarian clear cell carcinoma and validated its effects in both 2D and 3D bioprinted OCCC models. The paper concluded that ibrutinib demonstrates potential for the treatment of certain rare subtypes of ovarian cancer and should be further investigated.
In the 2026 ivacaftor study, the group worked collaboratively to test ivacaftor in ROR1-expressing high-grade serous ovarian cancer cell lines, 3D bioprinted models, and patient-derived organoids. The study reported significant anti-tumour potential in preclinical HGSOC models, supporting further investigation of ivacaftor as a repurposed therapy for ovarian cancer.
Together, these papers show the significance of Amy’s contribution. She is helping extend an established platform and workflow into new compounds, new models, and new ovarian cancer questions.
A Stronger Bridge to Precision Oncology
Amy already sees where this work is heading next.
“We’re about to optimise bioprinting for patient-derived organoids,” she says. “Ideally, this could look like a semi-automated approach to precision oncology, where we can test repurposed or novel drug candidates on 3D cultures that have been generated directly from patient tissue.”
Deborah points to another active area of development as well: 3D migration models designed to better mimic cell movement in 3D structures than a traditional 2D wound-healing assay.
Taken together, these efforts point toward a broader vision for RASTRUM in translational oncology. Not as a replacement for every model, but as a reproducible, scalable, and biologically relevant platform that helps researchers make better decisions before moving into the clinic.
From Better Models to Better Decisions
For Amy, the value of NAMs in ovarian cancer is not abstract. It is practical, urgent, and deeply connected to patient need.
“It is a genuinely exciting space to be working in right now,” she says.
That excitement is justified. Ovarian cancer remains a difficult disease with significant unmet need. Drug repurposing offers a faster path to new options, but only if the evidence supporting those choices is strong enough to matter.
This is where RASTRUM fits so well in the Marsh lab’s workflow. It gives the team a reproducible, biologically relevant way to validate candidates after in silico discovery and before precious patient material is used at scale. It helps uncover responses that 2D can miss. And across both the earlier efavirenz work and the newer studies Amy is helping drive, it has become part of a broader translational strategy for making stronger decisions in ovarian cancer drug development.
That is what makes this work so significant. It shows how stronger preclinical models can do more than improve data quality. They can help researchers make better decisions about which therapies deserve to move forward.