Neck masses, especially in the setting of previous treatment for head and neck cancer, can e very complex to interpret. In the past, there has been no standardization of optimal upfront imaging, surveillance algorithms, or management recommendations. In response, the ACR formed a committee with two goals:
1. Produce a lexicon to describe neck masses on CT, PET, and MRI.
2. Based on this lexicon, develop a standardized risk stratification system to inform practitioners about which neck masses warrant biopsy.
The result was the ACR Neck Imaging Reporting & Data System (NI-RADS™), which is aimed at improvement in these aspects. The Bulletin spoke with Amy F. Juliano, MD, chair of the ACR NI-RADS Committee and a head and neck radiologist in the department of radiology at Massachusetts Eye and Ear/Harvard Medical School, to learn more about the importance of standardizing imaging surveillance and reporting and how radiologists can use the ACR NI-RADS to improve care for head and neck cancer patients.
Neck masses, especially in the setting of previous treatment for head and neck cancer, can be very complex to interpret. Can you give us some context for the challenges and what led to the development of NI-RADS?
Imaging surveillance after treatment for head and neck cancer is challenging because of the complexity and varied types of resection and reconstruction surgeries, in addition to post-treatment changes from radiation and chemotherapy. Tumors can be at various anatomic sites, each associated with different risk factors, staged differently, and carrying different prognostic and treatment implications. Post-treatment surveillance head and neck imaging studies are even harder to interpret. The picture is often complicated further given the distorted post-treatment appearance, varying surgical techniques and types of flaps that can be used — and with radiation and chemotherapy often complicating the picture further.
For those unfamiliar with head and neck imaging — and even for those who are — reporting on these scans can be challenging and time-consuming. What’s more, the reporting can vary widely — whether in terms of the language used, impression rendered (e.g., Is there recurrence or not? What is the degree of suspicion?), and management recommended. All of this can complicate the overall clarity of the report, which ultimately determines user friendliness to the patient, referring clinicians, and other radiologists interpreting subsequent surveillance scans.
What is NI-RADS and why is it an indispensable part of quality and safety?
NI-RADS offers a widely applicable, understandable, and validated template for the management of neck masses on the basis of CT, PET, and MRI features. The end goal is to increase the quality and utility of our radiology reports, increase consistency, decrease inter-observer report variability, and provide discrete imaging conclusions that are clinically useful, consistent, and actionable.
The NI-RADS template guides standardized reporting on the basis of this lexicon with defined levels of suspicion and management recommendations. It also helps radiologists provide guidance regarding the management of patients after treatment for head and neck cancer.
At a most basic level, the NI-RADS template provides structure for our though processes, tells us what we should be looking for and how to decide whether what we see is concerning or not, and helps put thoughts into words on our reports when it comes time to dictate.
What are the benefits of using NI-RADS for reporting head and neck cancer?
NI-RADS helps radiologists differentiate benign from malignant findings, offers an algorithm to stratify risk, and enables clear and concise communication to referring clinicians to better guide clinical management (e.g., routine surveillance vs. short-term follow-up vs. biopsy). The template is simple, clear, and easy to adopt clinically for everyday use.
Among those who have implemented NI-RADS, the feedback has been great. The post-treatment neck is often a source of anxiety for diagnostic radiologists. Radiologists and trainees using NI-RADS often find that reading these scans is much less intimidating. At a most basic level, the NI-RADS template provides structure for our thought processes, tells us what we should be looking for and how to decide whether what we see is concerning or not, and helps put thoughts into words on our reports when it comes time to dictate.
Beyond that, a standardized template also produces reports that facilitate future data mining, which can be used for research and refinement in diagnostic accuracy. This enables retrospective data mining and utilization of AI. We can go back and look at old reports and, armed with any available histologic correlation or knowledge of long-term outcome, we can gauge accuracy of our imaging interpretation, pinpoint where and how discrepancies occur, and make improvements to provide better patient care.
What is the radiologist’s role in enhancing care for head and neck cancer patients?
We can offer guidance for the best imaging modality to address pertinent clinical questions, produce accurate and consistent interpretation, and perform image-guided biopsies. Radiologists play an integral role in reading room consults, tumor boards and interdisciplinary clinics/conferences. Some organizations even embed radiologists directly within the head and neck surgery clinic area, where radiologists have live interactions and discussions with the surgeons and oncologists — and at times, patients — during the clinic visit/imaging episode.
During these interactions, the conversation is clearer and more concise when the NI-RADS template and mindset is used. Everyone can more easily get on the same page when using consistent language, and there can be a more organized and quantum assessment of suspicion for tumor recurrence. The report template leads to two discrete numbers that assign risk/level of suspicion for tumor recurrence for the primary site and neck/nodes. The NI-RADS numbers link to management guidelines, so it is clear to clinicians what our recommendations are based on imaging findings.
What’s next for NI-RADS and what excites you about the future?
I’m excited about the new NI-RADS table for MRI. It just got rolled out and not too many people know about it yet. We’re also looking into conducting research to answer critical questions like these:
• Do radiologists agree with each other in interpretation using NI-RADS?
• Do the assigned NI-RADS category numbers match the findings descriptors we use?
• How can we add AI to the mix to assess accuracy and improve?
• How can practices can get started with NI-RADS?
Visit the NI-RADS page on the ACR website where you can find easy-to-use templates as well as training modules and webinars. Watch for NI-RADS lectures and workshops at various society meetings. We are always looking for more people from various practice settings, areas of expertise, and geographical locations to get involved. The more perspectives and ideas we have, the better we can improve what we already have — and be inspired about future directions that can best serve our patients and clinical colleagues.