Tumors are not monolithic masses but highly heterogeneous and spatially organized entities. Within these three-dimensional structures, a myriad of cell types, including malignant cells, stromal elements, immune cells, and vascular components, co-exist and interact in a tightly choreographed yet chaotic manner. This complex web of interactions governs the tumor’s behavior—its growth, progression, potential to invade surrounding tissues, and capability to metastasize. Historically, studies have examined tumors largely through dissociated cells or thin tissue sections, providing snapshots that fail to capture the holistic spatial context of tumor microenvironments and their evolution.

The emergence of spatial multi-omics technologies is revolutionizing this landscape by integrating genomic, transcriptomic, proteomic, and metabolomic data with spatial resolution. By preserving the architectural integrity of tumor tissues, scientists can now map molecular profiles directly onto three-dimensional landscapes. This progression is pivotal because cellular function and fate are often dictated not merely by intrinsic properties but by their spatial context and interaction with neighboring cells and extracellular matrices. The ability to visualize where, when, and how molecular signals propagate within tumors offers unprecedented insights into cancer biology that were previously inaccessible.

Creating 3D tumor atlases entails the integration of these spatially resolved multi-omics data, producing comprehensive maps that delineate tumor cell populations, stromal niches, vascular networks, and immune infiltrates within intact tissue volumes. Such atlases are dynamic, capable of capturing temporal changes across tumor initiation, progression, and metastasis. They enable researchers to track the evolutionary trajectories of cancer cells and their interactions with the microenvironment over time, thus shedding light on the operational principles that govern tumor heterogeneity and adaptation.

An extraordinary challenge in this domain is the sheer scale and complexity of the data generated. Sophisticated computational tools and machine learning algorithms are indispensable for data integration, visualization, and interpretation. These technologies facilitate the reconstruction of high-resolution 3D tumor models and the identification of spatially restricted molecular signatures that could serve as novel biomarkers. Furthermore, this computational prowess enables the dissection of intricate cellular crosstalk, revealing potential vulnerabilities in tumor ecosystems that might be exploited therapeutically.

Among the promising applications of 3D tumor atlases is their role in risk stratification and early cancer detection. By capturing precancerous lesions and the initial molecular changes that precede overt malignancy, these atlases could transform screening practices. Early interventions informed by precise molecular maps may prevent disease progression or enable more effective, less invasive therapeutic strategies, remarkably improving patient outcomes. This proactive approach represents a paradigm shift from reactive treatment to preemptive cancer management.

The tumor microenvironment is another critical aspect illuminated by 3D atlases. Immune cells infiltrate tumors in heterogeneous patterns, with spatial distributions affecting immune evasion and responses to immunotherapy. Mapping these spatial immune landscapes at high resolution allows for a better understanding of immunological “cold” and “hot” tumors, thereby guiding the design and optimization of immunotherapeutic regimens. As immunotherapies become increasingly central to oncology, spatial multi-omics provides a valuable framework for personalizing treatment.

Beyond immune cells, cancer-associated fibroblasts (CAFs) and other stromal components play multifaceted roles in tumor progression and therapy resistance. The structural and functional mapping of CAF subpopulations unveils their diverse contributions within tumor niches. Three-dimensional atlases facilitate the spatial localization of these subpopulations alongside tumor cells, revealing patterns of influence on tumor architecture and therapy responses. Targeting specific stromal components identified in spatial contexts could enhance therapeutic efficacy and overcome resistance mechanisms.

Metastasis—the deadly hallmark of cancer—also gains new investigative tools through 3D spatial omics. By charting the molecular evolution and spatial dissemination of metastatic clones from primary tumors across multiple sites, these atlases delineate the trajectories and mechanisms of cancer spread. Understanding how metastatic niches establish and thrive within distinct tissue microenvironments opens possibilities for intercepting metastasis at early stages, potentially reducing mortality rates associated with late-stage cancer.

The construction of these atlases is bolstered by novel technological platforms, including high-resolution imaging mass cytometry, spatial transcriptomics, and multiplexed immunohistochemistry. These approaches permit the simultaneous assessment of tens to hundreds of molecular markers in situ, preserving spatial contexts at single-cell or subcellular resolutions. Integration of these data types into 3D frameworks requires harmonization of disparate datasets and stringent quality controls to ensure biological validity. Interdisciplinary collaborations among biologists, engineers, and data scientists are therefore crucial to pushing the frontiers of this field.

As these technological horizons expand, so do the challenges associated with clinical translation. Incorporating spatial multi-omics into routine diagnostics involves scaling these complex assays, reducing costs, and ensuring reproducibility and clinical relevance. Robust computational pipelines capable of delivering actionable insights within clinically acceptable timelines are essential. Furthermore, ethical considerations regarding patient data privacy and consent for extensive molecular profiling remain paramount and warrant diligent attention.

The potential impact of 3D multi-omics tumor atlases extends beyond immediate clinical applications, offering new avenues for fundamental cancer research. By providing a spatially resolved molecular atlas of tumor ecosystems, researchers can investigate the fundamental mechanisms driving tumor heterogeneity and resistance evolution. Such insights can unveil novel therapeutic targets that disrupt critical tumor-microenvironment interactions, ultimately fostering innovative drug development strategies.

In sum, the advent of 3D multi-omics tumor atlases represents a transformative leap forward in oncology, bridging the gap between molecular detail and spatial context across tumor ecosystems. These atlases integrate high-dimensional data across multiple scales, from molecular to cellular to tissue architectures, and capture temporal tumor dynamics in unprecedented detail. Their capacity to elucidate the complexity of tumor biology promises revolutionary advances in early detection, personalized therapy, and ultimately, cancer prevention.

As this field continues to unfold, the synergy of cutting-edge technologies, computational innovations, and clinical aspirations will shape a future where cancer interception becomes both precise and proactive. The path forward entails refining atlas generation, enhancing accessibility, and fostering collaborative networks that accelerate translation from bench to bedside. This holistic approach, empowered by spatial multi-omics, may finally tip the scales in favor of patients in the ongoing war against cancer.

Subject of Research: Development and application of three-dimensional spatial multi-omics tumor atlases to understand tumor heterogeneity, evolution, and clinical translation.

Article Title: 3D multi-omics tumour atlases: from technology to biology and clinical translation.

Article References:
Liu, M., Villazon, J., Forjaz, A. et al. 3D multi-omics tumour atlases: from technology to biology and clinical translation. Nat Rev Cancer (2026). https://doi.org/10.1038/s41568-026-00940-0

Image Credits: AI Generated

image: Figure 2: Maximum-intensity projection images from [⁶⁸Ga]Ga-PSMA and [¹⁸F]-FDG PET before treatment, and [¹⁷⁷Lu]Lu SPECT displayed without and with (red) segmentation with SUV of >3 in 74-y-old mCRPC patient. [¹⁷⁷Lu]Lu-PSMA therapy was stopped at third injection despite dramatic decrease in PSA because of lack of favorable impact on symptoms and worsening clinical condition, and patient died 6 months later. High TLA was observed after second and third injections with advent of new bone lesions (orange arrows), both being indicative of poor prognosis.

view more 

Credit: Image created by L Imbert and C Boursier, et al., Regional University Hospital Center in Nancy, France.

Reston, VA (October 13, 2025)–A new fast and convenient approach to scintigraphy-based monitoring allows physicians to efficiently and reliably assess prostate cancer progression or regression during treatment. With this strong prognostic information, treatments for prostate cancer patients can be personalized according to tumor evolution, significantly impacting their overall survival. This research was published online in The Journal of Nuclear Medicine.

¹⁷⁷Lu-PSMA is a highly-targeted radionuclide treatment for metastatic castration-resistant prostate cancer (mCRPC). Conventional monitoring techniques during ¹⁷⁷Lu-PSMA therapy include analysis of the clinical condition and prostate-specific antigen (PSA) measurements. PSMA PET/CT is also performed before and during treatment. ¹⁷⁷Lu SPECT imaging after each injection has shown potential for treatment monitoring, however, it often has long imaging times.

“This drawback makes it challenging to systematically perform SPECT after each ¹⁷⁷Lu-PSMA injection including after the last injection, which may provide the most substantial prognostic information,”  said Laetitia Imbert, PhD, a physicist at Regional University Hospital Center in Nancy, France. “However, this technical issue can now be largely overcome with the advent of high-sensitivity 360 cadmium zinc telluride (CZT) SPECT systems, which enable total-body SPECT recordings in under 20 minutes. Our study sought to explore this new SPECT approach further in regard to ¹⁷⁷Lu PSMA treatment monitoring.”

The retrospective study included 72 mCRPC patients who received up to six ¹⁷⁷Lu-PSMA treatments. All patients underwent ⁶⁸Ga-PSMA-11 PET before initial treatment, had their PSA measured before each injection, and underwent ¹⁷⁷Lu-PSMA CZT SPECT after each injection. Quantitative image analysis and statistical image analysis were performed to predict overall survival.

Most PSA, PET, and CZT SPECT variables were significant univariate predictors of overall survival. However, only two ¹⁷⁷Lu CZT SPECT variables were multivariate predictors: detection of new bone lesions during treatment and final total lesion activity (TLA). Means of survival times were 19.7 months in the 19 patients who showed no new bone lesions and a low level of TLA; 14.4 months in the 19 patients with only one of these two criteria; and 6.9 months in the 19 patients with neither criterion. 

“Fighting cancer is a battle against time, as the disease can evolve rapidly. This new scintigraphy camera will assist in changing treatment plans for patients who do not respond well, at the earliest stage possible, noted Caroline Boursier, MD, a physician at Regional University Hospital Center in Nancy, France. Improvements could be achieved by adjusting the injected activity of ¹⁷⁷Lu-PSMA, substituting this radionuclide therapeutic agent with an alternative, or combining it with other cancer treatments such as chemotherapy, external radiotherapy, immunotherapy, or anti-angiogenic drugs.”

She added, “Logistically, monitoring patients with CZT SPECT is easy to schedule because of the very fast imaging times and the fact that no additional tracer injection is required. Scintigraphy imaging can therefore begin as soon as the patient arrives at the nuclear medicine department, and in our experience, they can leave in less than 30 minutes.”

The authors of Prognostic Value of Comprehensive Analysis of Metastatic Prostate Tumor Changes from First to Last [¹⁷⁷Lu]Lu-PSMA Therapy Injections Through Serial High-Speed Whole-Body 360 Cadmium Zinc Telluride SPECT include Julien Kunsch, Pierre Olivier, and Marine Claudin, Department of Nuclear Medicine, CHRU Nancy, Nancyclotep Imaging Platform, Nancy, France; Timoth e Zaragori, Innovation Technologique, CIC 1433, CHRU-Nancy, INSERM, Universit de Lorraine, Nancy, France, and IADI, INSERM, U1254, Universit de Lorraine, Nancy, France; Pierre-Yves Marie, S bastien Heyer, Antoine Verger, Laetitia Imbert, and Caroline Boursier, Department of Nuclear Medicine, CHRU Nancy, Nancyclotep Imaging Platform, Nancy, France, and IADI, INSERM, U1254, Universit de Lorraine, Nancy, France; and Perrine Raymond, Department of Endocrinology, CHRU Nancy, Nancy, France.

Please visit the SNMMI Media Center for more information about molecular imaging and precision imaging. To schedule an interview with the researchers, please contact Rebecca Maxey at (703) 652-6772 or rmaxey@snmmi.org.

About JNM and the Society of Nuclear Medicine and Molecular Imaging
The Journal of Nuclear Medicine (JNM) is the world s leading nuclear medicine, molecular imaging and theranostics journal, accessed 15 million times each year by practitioners around the globe, providing them with the information they need to advance this rapidly expanding field. Current and past issues of The Journal of Nuclear Medicine can be found online at 

JNM is published by the Society of Nuclear Medicine and Molecular Imaging (SNMMI), an international scientific and medical organization dedicated to advancing nuclear medicine, molecular imaging, and theranostics precision medicine that allows diagnosis and treatment to be tailored to individual patients in order to achieve the best possible outcomes. For more information, visit www.snmmi.org.

Journal

Journal of Nuclear Medicine

DOI

10.2967/jnumed.125.270358