New research suggests that the combination of MRI assessments for cellularity, intracellular volume fraction (Vin) and macromolecular proton fraction (MPF) provided more optimal differentiation of low-grade and high-grade cervical cancer than individual parameters.
For the prospective study, recently published in Academic Radiology, researchers reviewed data from 92 cervical cancer patients who had time-dependent diffusion MRI (Td-dMRI) and 3T macromolecular proton fraction (MPF) imaging. Fifty-seven patients had low-grade (grade I/II) cervical cancer, and 35 patients had high-grade (grade III) cervical cancer, according to the study.
For differentiating between low-grade and high-grade cervical cancer, the study authors found that the combination of the Td-dMRI parameters cellularity and Vin with MPF evaluation achieved a 96 percent AUC, 97.14 percent sensitivity and 82.46 percent specificity.
In contrast, the individual parameters had significantly lower AUCs, sensitivity and specificity. The researchers noted that cellularity offered a 92.3 percent AUC, 88.57 percent sensitivity and 77.19 percent specificity. The individual Vin assessment provided over a 10 percent lower AUC (85.4 percent), over 14 percent lower sensitivity (82.86 percent) and nearly 9 percent lower specificity (73.68 percent) in comparison to the aforementioned combination of parameters. Assessment of MPF also offered lower AUC (84.8 percent), sensitivity (91.43 percent) and specificity (68.42 percent) in contrast to the combination of parameters for differentiating cervical cancer, according to the study authors.
The researchers maintained that the Td-dMRI-derived parameters cellularity and Vin played a significant role in non-invasive grading of cervical cancer.
“An important explanation for this outcome may be the greater aggressiveness associated with high-grade lesions. Specifically, highly aggressive lesions exhibit enhanced cell proliferation—a process that directly induces an increase in cell density and intracellular volume fraction, while simultaneously reducing extracellular space. Further, incomplete cellular maturation and intercellular compression, driven by this rapid proliferation, also lead to a decrease in cell diameter. Collectively, these pathological changes ultimately result in elevated cellularity and Vin, as well as reduced diameter and Dex, (extracellular diffusivity)” noted lead study author Nan Meng, MD, who is affiliated with the Department of Radiology at Henan Provincial People’s Hospital and Zhengzhou University People’s Hospital in Zhengzhou, China, and colleagues.
Three Key Takeaways
• Multiparametric MRI markedly improves tumor grading accuracy. The combined use of cellularity, intracellular volume fraction (Vin), and macromolecular proton fraction (MPF) achieved superior diagnostic performance (AUC 96 percent, sensitivity 97.1 percent, specificity 82.5 percent) compared to any single parameter, supporting a multiparametric approach for preoperative risk stratification.
• Td-dMRI metrics reflect underlying tumor biology. Elevated cellularity and Vin in high-grade tumors correspond to increased cell density, proliferation, and reduced extracellular space — key histopathologic features that can be non-invasively captured and leveraged for tumor grading.
• MPF adds independent and complementary value. Higher MPF values in high-grade disease — potentially reflecting collagen deposition and stromal changes — serve as an independent predictor, reinforcing the added value of integrating MPF with diffusion-based metrics for more robust differentiation.
The study authors noted that quantitative MPF was also an independent predictor for differentiating between high-grade and low-grade cervical cancer.
“Results showed that MPF values in the high-grade group were significantly higher than those in the low-grade group … We hypothesize that differences in collagen deposition across CC with varying malignancy grades may serve as a primary factor contributing to these observed MPF alterations. Additionally, tumor stromal proliferation and cellular death may also play a role in driving this change,” added Meng and colleagues.
(Editor’s note: For related content, see “Ten Takeaways from New Guidelines on Pre-Op MRI for Fertility-Sparing Treatments of Cervical, Endometrial and Ovarian Cancer,” “Can MRI Have an Impact with Fertility-Sparing Treatments for Endometrial and Cervical Cancers?” and “Early-Stage Cervical Cancer: Is MRI, CT or PET/CT the Best Option for Diagnosing Lymph Node Metastases?”)
Beyond the inherent limitations of a single-center retrospective trial, the authors acknowledged a relatively small cohort size, the lack of external validation and the exclusion of < 1 cm lesions due to spatial resolution constraints.
