Could measuring a tumor’s volume add useful information to assessments of cancer treatment? A 2026 study in European Radiology explored that question in patients with colorectal cancer that had spread to the liver. Its findings suggest that AI-assisted volume measurements could complement established assessment methods, particularly when metastatic disease is confined to the liver.
For Dr. Colin Paul Spears, a medical oncologist and founder of Fibons, this research connects with a longstanding interest in how tumor measurements are interpreted. He believes studying three-dimensional change alongside conventional measurements could help researchers better understand differences in how individual lesions respond.
Spears recognizes the practical value of established methods. “From a practical standpoint, it’s reproducible,” he says of diameter-based assessment. His interest lies in exploring whether additional measurements could provide useful context while preserving the consistency that clinical research requires.
RECIST 1.1, the Response Evaluation Criteria in Solid Tumors, provides a standardized framework for assessing treatment response. It tracks sums of selected lesions’ diameters alongside other disease findings, including new lesions. Target lesions are measured along their longest diameter.
Spears’ concern is that the combined sums approach leaves some differences between individual lesions less visible. In his view, examining their estimated volumes could help researchers explore patterns of growth or shrinkage that a single percentage change does not fully convey.
“Comparing tumors means considering both their starting size and how they change during treatment,” he says. “This could give us a clearer understanding of how each tumor and summed parameter volumes respond.”
His work on this subject includes a 1984 paper in Medical and Pediatric Oncology examining volume-doubling rulers for estimating tumor volume from medical images. The paper also addressed geometric limitations, including how elongated shapes affect estimates. That work provides a mathematical foundation for his continuing interest in measurement, while the clinical usefulness of his current approach remains a question for further study.
Through Fibons, Spears has developed a physical ruler designed to convert diameter measurements into volume and area doubling numbers. He explains that each additional volume-doubling step represents twice the estimated volume. The approach uses geometric relationships to express size differently, with assumptions about lesion shape remaining important.
Spears also envisions software that could organize those measurements and compare response thresholds. “Once the radiologist assigns a doubling number to the tumor measurement, a computer can handle the remaining calculations,” he explains. “The aim is to provide additional information without substantially increasing the radiologist’s workload.” He hopes to examine that potential advantage through clinical research.
The European Radiology study offers a separate example of volumetric assessment under investigation. Researchers assessed 157 patients using both RECIST 1.1 and an AI-based volumetric method. Both approaches provided information associated with overall survival across the full cohort. Among 43 patients with liver-only metastases, volumetric assessment significantly distinguished survival outcomes, while RECIST did not reach statistical significance. These findings concern the study’s AI method and do not establish the performance of Spears’ ruler.
For Spears, the next step is comparison in suitable patient populations. He identifies lung tumors and colorectal cancer lesions in the liver as areas of interest. He also hopes to investigate tumors with inactive central regions and explore how volume estimates relate to other indicators of treatment response.
Consistency remains an essential question. He says, “We still need to establish whether the method produces consistent results when measurements are repeated by the same person or taken by different people.” Spears wants to examine both the reliability of the approach and how readily computer algorithms could support it.
He uses the term “volume of involved tissue” to explain his interest in the broader tissue affected by a tumor. In his proposed framework, interpreting changes depends on assumptions about whether a lesion retains a comparable shape over time. The response criteria for lymphomas follow the Lugano criteria that use the sum of perpendicular diameters of lesions. This approach is also called the Cheson criteria and is also used by the World Health Organization. The Lugano approach is essentially a cross-sectional area method of tumor assessment for response and progression.
Spears’ doubling ruler approach harmonizes RECIST and Lugano criteria tumor response because three doublings of volume equal two doublings of area. He points out the 1-2-3-4 of doubling. That is, one doubling of radius equals two doublings of area, equals three doublings of volume, and four times the cross-sectional area equals the surface area. He also points out that the doubling approach is applicable to irregular lesions that grow or shrink as long as they maintain the same general shape and orientation.
Spears’ doubling number approach is readily adapted to electronic tracking of lesions’ size for data enrichment, for a better comparison of treatment arms and choosing different thresholds for determination of progression. In other words, the doubling approach promises to harmonize all the major systems of tumor response.
Ultimately, Spears hopes this work can help researchers compare response criteria within a shared framework. He favors studying volumetric information alongside established assessments to determine where it adds value. As AI-assisted imaging research develops, he sees an opportunity to investigate whether additional measurements can make patterns of treatment response clearer and more useful to those interpreting them.
