Editorial for "In Vivo Microstructure Imaging in Oropharyngeal Squamous Cell Carcinoma Using the Random Walk With Barriers Model"

Cell size, a commonly assessed cytological characteristic, may play an important role in cellular functions, and measurement thereof could shed light on specific alterations occurring within biological tissues.1 Cancer cells can change their size to survive drug treatment or other challenges within their environment.2 Thus, studying cell size can be useful in clinical practice for determining the efficacy of treatments of malignant tumors. Several techniques are available to measure cell size, including in vivo imaging and microscopy. Novel magnetic resonance imaging (MRI) methods, such as oscillating-gradient spin-echo, can shorten the diffusion time, which aids microlevel evaluation of water molecules. However, there are some limitations to in vivo measurements, and clinical application still needs to be improved. Cao et al. explored a method that could be used in clinical practice.3 Cell size measurements obtained by microscopy and oscillating-gradient spin-echo were compared to evaluate treatment efficacy for squamous cell carcinoma. One of the unique advantages of diffusion MRI is that it provides information at the micrometer scale in vivo, whereas diffusion-weighted imaging data are of millimeter scale. A water molecule's apparent diffusion distance, <χ2> 1/2, can be calculated from the apparent diffusion coefficient (ADC) using Einstein's mean quadratic displacement equation. MR imaging with different diffusion times has become possible due to hardware advances in MRI gradient systems. In an environment with no structure, such as free water (Gaussian diffusion), the motion of water molecules exhibits the same diffusion coefficient regardless of the diffusion time. However, in an environment with some degree of barrier structure (inhibited or restricted diffusion), everytime a water molecule encounters a barrier it exhibits a different diffusion coefficient. As a result, the diffusion coefficient and distance are shorter than in a free-diffusion environment. The ADC has shown some value in the investigation of tumors at the μm scale. In addition, numerous models have been proposed, such as imaging microstructural parameters using limited spectrally edited diffusion (IMPULSED), pulsed and oscillating gradient MRI for assessment of cell size and extracellular space (POMACE), and vascular, extracellular, and restricted diffusion for cytometry in tumors (VERDICT),3 and the authors have employed the random walk with barriers model. This model differs from others in taking into account the impact of permeable cellular membranes allowing the diffusion of water molecules. One study found that the V/S values of tumors were significantly correlated with clinical stage, exhibiting an increase as clinical stage progressed. However, this needs to be considered in the context of the population of that study; the patients had stage I–III tumors with p16 positivity and stage IV tumors with p16 negativity, where human papillomavirus (HPV)− tumors have higher ADC values than HPV+ tumors in oropharyngeal squamous cell carcinoma4; many HPV-related oropharyngeal cancers are p16-positive. In addition, analysis of early cellular responses in tumors demonstrated a significant increase in D0, while nonsignificant increases were observed in κ and V/S. An increase in the diffusion coefficient of tumors under treatment has been shown in various cancers, including of the head and neck.5 Some changes in κ and V/S might merit investigation beyond ADC; studies including more patients and scans throughout the treatment course are needed. Changes in diffusion contrast, particularly when transitioning from short to long diffusion times, is an intriguing area of research with the potential to shed light on various aspects of interest, including the ADC. However, it is important to acknowledge that such investigations often require modelization, like in this study. Despite their potential, it is important to recognize that the assumptions underlying the models may not always hold true, particularly when dealing with different diffusion times. As evidenced by a recent publication,6 the assumptions of such models can be called into question. Validation of these models is vital, but is not without challenges. Determining how to accurately assess permeability remains a hurdle that researchers need to address to advance this field. Additionally, the diffusion time is not displayed on the console, which makes it necessary to report the diffusion time in clinical studies. Despite the small size of the sample and the heterogeneity of the lesion characteristics, the proposed diffusion MRI model successfully provided valuable insights into microstructural features of head and neck squamous cell carcinoma. This approach has the potential to serve as a virtual microscope, or even as a noninvasive alternative to biopsies, allowing for the investigation of cancer microstructure at the cellular level. However, simple ADC, which is very robust and model free, is sensitive in principle to microstructure. This is because the diffusion time used in practice on clinical scanners is long enough to allow water molecules to explore tissue features at a microscopic scale. Further development of the MRI-based cell size measurement method by incorporating other methods is desirable.1 We thank Michael Irvine, PhD, from Edanz (https://jp.edanz.com/ac) for editing a draft of this manuscript.