Lymphoma accounts for the majority of hematopoietic malignancies in dogs, and more than 80 percent of cases present as multicentric disease, with tumors distributed across multiple lymph nodes. Accurate staging matters: knowing whether the spleen, liver, or bone marrow is involved determines the World Health Organization stage, shapes prognosis, and guides treatment decisions. Yet the standard approach is cumbersome. Veterinarians typically combine a physical examination, complete blood count with blood smear analysis, serum biochemistry, three-view thoracic radiographs, abdominal ultrasound, and cytologic evaluation of the liver, spleen, and bone marrow. Each test carries its own costs, procedural risks, and dependence on operator skill, and inconsistencies between protocols produce what researchers call stage migration, complicating comparisons between clinical studies.

In human medicine, the gold standard for lymphoma staging is positron emission tomography combined with computed tomography, PET/CT, a functional imaging technique that reveals metabolically active tumor tissue throughout the body. But PET/CT has drawbacks even in humans: up to 30 percent of non-Hodgkin lymphoma cases show low or absent tracer uptake, indolent subtypes may escape detection, post-treatment inflammation can trigger false positives, and repeated scans deliver cumulative ionizing radiation, a concern highlighted by recent evidence linking diagnostic CT exposure in children to a mildly increased incidence of hematopoietic cancers. In veterinary medicine, PET/CT is largely inaccessible due to high costs and limited availability, leaving an obvious gap that whole-body diffusion-weighted MRI may be able to fill.

DWI exploits a subtle physical phenomenon: the random Brownian motion of water molecules within tissue. By applying magnetic field gradients, the technique measures how freely water diffuses, which in turn reflects cellular density and tissue architecture. Tumors, packed with densely clustered cells, restrict water movement and appear bright on diffusion-weighted images with corresponding dark signal on apparent diffusion coefficient, or ADC, maps. The Lyon team acquired images on a 1.5 Tesla scanner in three overlapping fields of view covering head, thorax, and abdomen, using single-shot echo-planar sequences at five different b-values ranging from 0 to 1400 s/mm². Merging the 3D T1-weighted Dixon images with diffusion-weighted and ADC maps via alpha-blending post-processing produced whole-body images that allowed a single 77-minute anesthetized session to survey lymph nodes, liver, spleen, and bone marrow simultaneously.

Five dogs with treatment-naïve, cytologically or histologically confirmed multicentric high-grade lymphoma were enrolled prospectively between September 2022 and December 2023, spanning breeds from German Shepherd to Jack Russell Terrier with a median age of nine years. Each animal underwent conventional comprehensive staging and WB-DWI at diagnosis, then received a month of induction chemotherapy with either L-COP or L-CHOP before the entire evaluation was repeated. A board-certified veterinary radiologist, blinded to the conventional results and to whether a scan was taken before or after treatment, interpreted every MRI. A lesion was called positive if it showed hyperintense signal on diffusion-weighted images with matching hypointensity on ADC maps.

The results were encouraging but imperfect. Conventional methods classified three dogs as stage IV and two as stage V due to bone marrow infiltration. WB-DWI staged one dog as stage III, three as stage IV, and one as stage V, agreeing with conventional staging in three of five cases and achieving 72 percent overall agreement across all anatomical sites. Cohen’s Kappa statistics indicated moderate agreement for peripheral lymph nodes (κ = 0.531) and bone marrow (κ = 0.545), but only slight agreement for liver and spleen (κ = 0.0). Notably, WB-DWI failed to detect liver involvement in all five patients, despite cytology confirming hepatic infiltration in each case.

In some respects, the MRI outperformed the traditional toolkit. Enlarged lymph nodes were readily distinguishable by their marked hyperintensity, and the technique exposed retropharyngeal involvement in four patients and inguinal involvement in one that physical examination had missed. Retropharyngeal nodes can be hard to palpate, and inguinal nodes are often masked by adjacent mammary tissue or fat. The modality also seemed advantageous for intrathoracic and cranial abdominal node beds, which radiographs and ultrasound frequently obscure due to superimposition, small size, or gastrointestinal contents. Conversely, mesenteric lymph nodes proved difficult to evaluate, likely because of signal interference from adjacent intestinal loops, and a microchip artifact blocked assessment of one dog’s superficial cervical nodes. Most challenging of all were the liver and spleen, organs known in human studies to be susceptible to respiratory and cardiac motion artifacts and to display inherently high signal on diffusion-weighted sequences even when healthy.

The treatment-monitoring results were striking. By the end of induction chemotherapy, all five dogs had achieved complete clinical remission by conventional criteria, and WB-DWI independently confirmed the resolution of lymph node and bone marrow disease in every animal, with post-treatment scans showing no sites of abnormal hyperintensity. The two methods were completely concordant for post-treatment follow-up. WB-DWI also flagged bone marrow infiltration in one of the two cytologically positive patients; the detected case had 20 percent marrow involvement versus 8 percent in the missed case, hinting that sensitivity depends on the degree of infiltration. Because bone marrow aspirates are not consistently performed in veterinary staging despite the prognostic weight of stage V disease, a non-invasive whole-body surrogate could meaningfully change practice.

The authors are careful to emphasize the study’s limitations. With only five dogs, limited stage variability, and no confirmed non-lymphoid organ involvement, the cohort cannot establish true diagnostic accuracy. Diffusion hyperintensity is not specific for cancer: increased cellularity, inflammation, or ischemia can all restrict water diffusion, and normal spleen and lymph nodes may share imaging characteristics with diffuse neoplastic infiltration. Quantitative ADC analysis was not performed because standardized reference ranges do not yet exist for dogs, although small studies have begun characterizing ADC values in healthy canine organs. Future work incorporating healthy controls, patients with inflammatory diseases, and PET/CT as a comparative reference standard will be essential to define the modality’s specificity and reliability across institutions.

Even so, the proof of concept lands at an interesting moment for comparative oncology. Canine multicentric lymphoma shares clinical and pathological features with human non-Hodgkin lymphoma, and advances in veterinary imaging often mirror and inform human practice. Human studies already report excellent agreement between WB-DWI and PET/CT for lymphoma staging, and a whole-body diffusion MRI atlas of healthy adults now provides benchmarks for quantitative interpretation. Building an equivalent canine atlas, refining protocols with additional sequences such as T2-weighted imaging, and validating findings in larger and more diverse populations are the logical next steps. If those efforts succeed, a single, radiation-free MRI examination could transform how veterinary oncologists stage and monitor one of the most common cancers in dogs, sparing patients multiple invasive procedures while delivering a more complete picture of disease in one session.

Subject of Research: Whole-body diffusion-weighted MRI for staging canine multicentric high-grade lymphoma

Article Title: Evaluation of whole-body diffusion-weighted magnetic resonance imaging for the staging of canine multicentric high-grade lymphoma: a proof-of-concept study

Article References: Barrett, L. E., Ségard-Weisse, E., Girard, R., Bernardo Marques, G., Chamel, G., Diop, N., Dimeglio, F., Lajoinie, M., Lincet, P., Eddin, J. N., Bandiera-Mezei, L., Genieis, D., Dorez, H., & Ponce, F. (2026). Evaluation of whole-body diffusion-weighted magnetic resonance imaging for the staging of canine multicentric high-grade lymphoma: a proof-of-concept study. Veterinary Oncology, 3(1), Article 7. https://doi.org/10.1186/s44356-026-00059-2

Image Credits: AI Generated

DOI: 10.1186/s44356-026-00059-2

Keywords: canine lymphoma, whole-body diffusion-weighted imaging, WB-DWI, cancer staging, veterinary oncology, magnetic resonance imaging, apparent diffusion coefficient, bone marrow infiltration, chemotherapy response, L-CHOP, non-invasive imaging, stage migration