The development of Reporting and Data System (RADS) began as an initiative by the
American College of Radiology (ACR) to decrease interobserver report variability among
radiologists and to deliver care that is “patient-centric, data-driven, and outcomes-based.”[1] RADS are mainly guidelines for lesion identification, characterization, and structured
reporting.
Interpretation of follow-up imaging of patients with head and neck squamous cell carcinomas
poses a great challenge. Patients undergo extensive surgical resections with composite
free-flap reconstructions, altering normal anatomy, and making radiologic interpretation
difficult. Chemoradiation results in soft tissue swelling, edema, and contrast enhancement
from inflammation or granulation tissue, all of which may be mistaken for recurrent
tumor. The Neck Imaging Reporting and Data System (NI-RADS) is a guide to report on
surveillance of head and neck cancers after therapy. The categories of NI-RADS are
designated as negative, low suspicion, high suspicion, and definite recurrence. In
this issue, Kumar et al have analyzed the efficacy of NI-RADS rating using contrast-enhanced
computed tomography (CT) in predicting local and regional recurrence of malignancies
after chemoradiotherapy.[2] This study appears to prove that enhanced CT would suffice in assigning NI-RADS
score when positron emission tomography/computed tomography (PET/CT) is not available.
The first RADS that came into being was the Breast Imaging Reporting and Data System
or BI-RADS, developed for mammography for the detection of breast cancer. Subsequently,
other systems followed, TI-RADS for thyroid nodule imaging, LI-RADS for chronic liver
disease and hepatocellular carcinoma imaging, Lung-RADS for lung cancer screening,
and PI-RADS for prostate cancer imaging. The list keeps growing—NI-RADS (neck), O-RADS
(ovarian), VI-RADS (bladder), MY-RADS (myeloma), Met-RADS (prostate), Onco-RADS, Node-RADS,
Bone-RADS, Coronary-RADS, CO-RADS (COVID-19) …..Whew!! Just as when we thought most
pathologies and organs were covered, we find a bevy is in the pipeline: PE-RADS (pulmonary
embolism), MSK-RADS (soft tissue), BT-RADS (brain tumor), KI-RADS (kidney masses),
Stroke-RADS…. And yet more in the offing.
Do we need them all? Are we pushing the RADS too far? When we have a plethora of RADS
across organs, modalities, and pathologies, don't these overwhelm the radiologists?
Are there incremental benefits to the patient or the clinician? Would the general
practitioner or family physician understand these fast-evolving terminologies? Have
we reached the point of diminishing gain? These are all points to ponder in the Indian
setting where radiology practices are more heterogeneous than many other countries.
The use of BI-RADS has caught on in India, but other RADS systems, perhaps except
for PI-RADS and LI-RADS, will take time to find their way into radiology reports consistently.
The use of classification or terminology, which is unfamiliar to clinicians, does
no help for patients. Clinicians have to be educated about each RADS system for such
telegraphic communication to be successful. Otherwise, we are talking into the wind.
Implementation might be easier at major institutions, but for these classifications
to diffuse towards smaller hospitals or rural practices, the going may be tough. The
ne plus ultra of communication might be directly discussing with the referring physician as to
what the nature of lesion would be, when to follow up, and if to biopsy or not. It
might be yet better if reports were discussed with patients such that they have a
better understanding of their condition, assuaging anxieties to a great extent.
Does the use of RADS make a report more communicative? Not really, if we consider
the wide variance in the BI-RADS category 4. Category 4b has a chance of cancer of
10 to 50% and 4c has a chance of malignancy of 50 to 95%. In terms of patient prognostication,
such a wide degree of variance is unacceptable.
RADS will keep evolving over time with inclusion of functional and quantitative imaging
that would be more accurate and predictive but at the same time more cumbersome for
radiologists. We have all seen that with the PI-RADS v1 and 2. Many authors feel PI-RADS
v1 is better than v2,[3] but some do not. We are left none the wiser.
Implementing RADS in regular practice across a country like India would require considerable
effort and training in standardization of protocols and structured reporting. More
importantly, we need to educate more than a million clinicians on how to interpret
this nomenclature. Can we have a simpler “Made in India” scoring system that will
be adopted more easily in our country? By decreasing variability in radiology reports
and clinical management, we can assuredly improve outcomes for patients.
I write this as one (of many) who is faced with the classic Hamletian dilemma—To RADS
or not to RADS. When we look at the situation objectively, RADS are here to stay,
but then, there is nothing like having discussions and multidisciplinary meetings
with our colleagues to communicate precisely and clearly and plan the best treatment
for a given patient. Even when a report with RADS score is sent out with an accurate
clinical interpretation we still have to debate, deliberate, and discuss, so is it
worth all the effort? Dictating is easy-putting a face on those words is far from
facile. I shall sign off leaving the readers to contemplate and decide what suits
them best in their practices.
Although our intellect always longs for clarity and certainty, our nature often finds
uncertainty fascinating.—Karl Von Clausewitz
