What Is a Category 4 Thyroid Nodule?

A Category 4 thyroid nodule is one rated “moderately suspicious” for cancer on the ACR TI-RADS scale, a widely used ultrasound scoring system that ranks thyroid nodules from 1 (benign) to 5 (highly suspicious). The designation is based on a points tally from specific features a radiologist sees on ultrasound, and it typically triggers a recommendation for biopsy once the nodule reaches a certain size. The label can be unsettling to see on a radiology report, but “moderately suspicious” does not mean cancer is likely or certain. The path from that label to an actual diagnosis involves several more steps, and most people end up with a benign result.

How the TI-RADS Scoring System Works

The American College of Radiology (ACR) published its Thyroid Imaging Reporting and Data System (TI-RADS) in 2017 as a standardized way for radiologists to evaluate thyroid nodules found on ultrasound. Rather than relying on gut feeling, the system assigns points across five feature categories: the nodule’s composition (solid, cystic, or mixed), its echogenicity (how bright or dark it appears compared to surrounding tissue), its shape, its margins, and whether it contains echogenic foci such as calcifications.¹

Each suspicious trait earns a set number of points. A completely solid nodule gets more points than a mixed or cystic one. A nodule darker than the surrounding strap muscles scores higher than one that is simply darker than normal thyroid tissue. An irregular or lobulated border adds points, as does a taller-than-wide shape. Microcalcifications, tiny bright specks within the nodule, carry the highest point value among echogenic foci, while rim or coarse calcifications carry fewer.¹

Once the points are added up, the nodule falls into one of five tiers:

TI-RADS 1: Benign (no points; a normal thyroid or simple cyst).
TI-RADS 2: Not suspicious (2 points).
TI-RADS 3: Mildly suspicious (3 points).
TI-RADS 4: Moderately suspicious (4–6 points).
TI-RADS 5: Highly suspicious (7 or more points).

A TI-RADS 4 nodule, then, has accumulated enough suspicious features to land in the middle-to-upper range of concern without maxing out the scale. It might be a solid, darker-than-normal nodule with one or two additional worrisome traits, but not the full cluster of features that would push it to a 5.

What Happens After a Category 4 Designation

A TI-RADS category does not, by itself, decide whether you need a biopsy. Size matters too. The ACR guidelines recommend fine-needle aspiration biopsy (FNAB) for TI-RADS 4 nodules when they measure 1.5 cm or larger. For comparison, TI-RADS 5 nodules have a lower size threshold of 1 cm, and TI-RADS 3 nodules are not flagged for biopsy until they reach 2.5 cm.¹

If your nodule falls below the size cutoff, your doctor will typically recommend periodic ultrasound monitoring rather than an immediate biopsy. If it meets or exceeds the threshold, a fine-needle aspiration is the standard next step. This procedure is done with a thin needle, usually guided by ultrasound, and generally takes only a few minutes. The cells collected are then examined under a microscope.

One study found that strictly following the TI-RADS size thresholds would have missed about 28% of cancers in nodules rated TI-RADS 3 through 5, because some malignant nodules were smaller than the recommended biopsy cutoff.² That same study found that malignancy rates were essentially the same regardless of nodule size, suggesting that the size threshold is a pragmatic compromise rather than a biological guarantee.² This is why some clinicians choose to biopsy smaller nodules when other clinical factors raise concern, like a family history of thyroid cancer or prior radiation exposure to the neck.

The Difference Between TI-RADS Category 4 and Bethesda Category IV

A common source of confusion is that two different “Category 4” systems exist in thyroid medicine, and they measure completely different things. TI-RADS Category 4 describes what a nodule looks like on ultrasound. Bethesda Category IV describes what the cells look like under a microscope after a biopsy. If you see “Category IV” on a cytology report rather than an imaging report, that is the Bethesda system, and it carries a different meaning.

The Bethesda System for Reporting Thyroid Cytopathology is a six-tier classification used by pathologists to report biopsy results. Bethesda Category IV is labeled “Follicular Neoplasm or Suspicious for a Follicular Neoplasm.” It indicates that the cells show features like microfollicle formation and some nuclear irregularity, but the pathologist cannot tell from a needle sample alone whether the growth is benign or malignant.³ To make that distinction, a surgeon often needs to remove part or all of the thyroid so a pathologist can examine the tissue architecture. This is the frustrating reality of follicular lesions: they cannot always be diagnosed definitively from a needle biopsy alone.

The malignancy rate for Bethesda IV nodules varies considerably across studies and institutions. One single-center study from a region with high thyroid cancer prevalence found a malignancy rate of about 47%, with papillary thyroid carcinoma being the most common cancer type and adenoma the most common benign finding.³ Other populations report lower rates. The wide range reflects differences in iodine intake, referral patterns, and how aggressively institutions biopsy nodules. In any case, the diagnosis is not settled by the Bethesda IV label alone. Further workup, whether molecular testing or surgery, is almost always needed.

Molecular Testing and How It Changes the Picture

For nodules that come back as Bethesda III (atypia of undetermined significance) or Bethesda IV (follicular neoplasm), molecular testing has become an increasingly common way to avoid unnecessary surgery. Two widely used platforms, Afirma GSC and ThyroSeq v3, analyze genetic markers from the biopsy sample to estimate how likely the nodule is to be cancerous.

A study of over 300 Bethesda III and IV nodules found that both tests had extremely high sensitivity, catching virtually all cancers. Afirma GSC had a sensitivity of 100% and ThyroSeq v3 came in at 97%. Their specificity, the ability to correctly identify benign nodules, was 77% and 83% respectively. The negative predictive value for both was at or near 100%, meaning that when these tests said a nodule was benign, it almost always was.⁴ This matters because a “benign” molecular result can spare you from a diagnostic surgery you might not need.

The clinical impact of molecular testing is measurable. A retrospective study found that in Bethesda IV nodules, about three-quarters of patients who underwent molecular testing received what researchers considered optimal surgical treatment, compared to roughly a third of those who did not have testing.⁵ “Optimal” here means the right scope of surgery on the first attempt: not too little (missing a cancer that requires a second operation) and not too much (removing the entire thyroid when only a lobe was needed). Molecular profiling can also predict tumor behavior, helping doctors tailor treatment beyond a simple benign-versus-malignant verdict.⁶

Molecular testing is not available everywhere and adds cost, so whether your doctor offers it depends on institutional resources and local practice patterns. Fine-needle aspiration remains the first-line diagnostic tool; molecular testing is an add-on for ambiguous biopsy results, not a replacement for the biopsy itself.⁷

When Cancer Is Found in These Nodules

Among Bethesda IV nodules that do turn out to be malignant, the follicular variant of papillary thyroid carcinoma (FVPTC) is consistently the most common cancer type across studies.⁸ A large dataset used to train an artificial intelligence diagnostic model found that among Bethesda IV nodules, roughly 47% were benign, about 33% were FVPTC, and around 20% were follicular thyroid carcinoma.⁹

Both follicular thyroid carcinoma and the follicular variant of papillary carcinoma tend to be slow-growing and carry a generally favorable prognosis when caught at a manageable stage. The more aggressive forms of thyroid cancer, like anaplastic or medullary carcinoma, are far less common in this category. So even when a Bethesda IV nodule does harbor cancer, the cancer itself is usually a treatable type.

Active Surveillance as an Alternative to Surgery

Not every Bethesda IV nodule needs to go straight to the operating room. Active surveillance, a strategy of careful monitoring with repeat imaging and sometimes repeat biopsy, is gaining ground for selected patients. A real-world study of patients with Bethesda IV nodules who were followed with active surveillance found high stability over time, low rates of malignancy progression, and low rates of eventual conversion to surgery, leading researchers to conclude that this approach is feasible for the right patients.¹⁰

Active surveillance is particularly appealing for patients who are older, have significant surgical risks, or have nodules with reassuring molecular test results. In resource-limited settings where molecular testing or specialized thyroid surgery is not easily accessible, ultrasound-guided monitoring can safely reduce unnecessary operations.¹⁰ That said, active surveillance requires commitment: regular ultrasound appointments, possible repeat biopsies, and the psychological weight of living with an uncertain diagnosis. It works best for patients who are comfortable with that uncertainty and whose medical team has a clear protocol for when to escalate to surgery.

Thermal Ablation for Small Bethesda IV Nodules

For patients who want to avoid traditional surgery but are not comfortable with simple monitoring, thermal ablation has emerged as a middle path. Techniques like radiofrequency ablation (RFA) and microwave ablation (MWA) use heat delivered through a needle-like probe to destroy nodule tissue without removing the thyroid.

A systematic review and meta-analysis of thermal ablation for Bethesda IV nodules found that nodule volume shrank by about 43% at one month, 69% at six months, and roughly 85% at twelve months. Both RFA and MWA achieved similar volume reduction rates at the one-year mark.¹¹ A separate study confirmed that RFA is effective for Bethesda IV nodules measuring 2 cm or less with lower-suspicion ultrasound features.¹²

Thermal ablation is not a universal replacement for surgery. It works best for smaller, lower-risk nodules and is not appropriate when there is strong suspicion of aggressive cancer or when the nodule has features suggesting it has spread beyond the thyroid. You will also need follow-up imaging afterward to confirm the nodule has not regrown. Still, for the right candidate, it offers a same-day outpatient procedure with faster recovery and preserved thyroid function.

Advanced Imaging Beyond Standard Ultrasound

Standard grayscale ultrasound is the workhorse for evaluating thyroid nodules, but it has limitations, especially for nodules that fall in the ambiguous middle tiers of TI-RADS. Researchers have explored add-on imaging techniques to improve diagnostic accuracy for Category 4 nodules.

Contrast-enhanced ultrasound (CEUS) uses microbubble contrast agents injected into a vein to visualize blood flow patterns within a nodule. Shear wave elastography (SWE) measures tissue stiffness, since cancerous tissue tends to be stiffer than benign tissue. Used individually, each technique performs reasonably well. But when CEUS and SWE are combined, sensitivity for detecting malignancy in Category 4 nodules jumped to about 95%, with accuracy exceeding 92%.¹³ The combined approach was significantly better at catching cancers than either technique alone, while maintaining similar specificity.

These advanced imaging tools are not yet standard at every hospital. They require specialized equipment and training, and insurance coverage varies. But for patients facing a diagnostic gray zone, particularly those with Category 4 nodules where biopsy results are inconclusive, these techniques can provide additional information that helps tip the decision toward or away from surgery.

Artificial Intelligence in Nodule Classification

One well-known problem with TI-RADS scoring is that different radiologists can look at the same ultrasound and arrive at different scores. Features like “lobulated margins” or “mildly hypoechoic” involve subjective judgment, and two experts may disagree on how many points a given nodule deserves. This interobserver variability means that the same nodule might be classified as TI-RADS 3 by one radiologist and TI-RADS 4 by another, changing the biopsy recommendation entirely.

AI-based decision support systems are being developed to address this. A clinical evaluation found that when radiologists used an AI tool alongside their own ultrasound readings, interobserver agreement on TI-RADS scores improved substantially, with the correlation rising from 0.74 to 0.98.¹⁴ That is a meaningful jump. Variability dropped and agreement with the established classification system increased. AI tools designed to classify Bethesda IV cytology slides are also under development, though these remain largely in the research phase.⁹

AI is not replacing radiologists or pathologists in thyroid nodule diagnosis anytime soon. But as a second set of eyes, particularly in settings where specialist radiologists are scarce, it has clear potential to reduce the kind of classification inconsistency that leads to patients being over- or under-biopsied.

Who Is More Likely to Have Thyroid Nodules in the First Place

Thyroid nodules are remarkably common. Most are discovered incidentally during imaging done for other reasons, and the vast majority are benign, with malignancy rates generally in the range of 7 to 15% across all nodules that undergo evaluation.⁷ Certain groups are more likely to develop them: women, people living in areas with iodine deficiency, and anyone with a history of radiation to the neck.¹⁵

Having a thyroid nodule, even one classified as TI-RADS 4, does not mean your thyroid is malfunctioning. Most people with nodules have normal thyroid hormone levels and no symptoms at all. The nodule is found, classified, and worked up based on imaging and biopsy findings, and for many people, the outcome is reassurance and a plan for periodic monitoring rather than any intervention. The classification systems exist precisely to separate the small number of worrisome nodules from the large majority that can be safely watched.

1
PubMed Central. Validation of TIRADS ACR Risk Assessment of Thyroid Nodules in Comparison to the ATA Guidelines
2
PubMed Central. The Size of a Thyroid Nodule with ACR TI-RADS Does Not Provide a Better Prediction of the Nature of the Nodule
3
PubMed Central. Malignancy rates in thyroid nodules classified as Bethesda categories III and IV; a subcontinent perspective
4
The Journal of Clinical Endocrinology & Metabolism. Bethesda III and IV Thyroid Nodules Managed Nonoperatively After Molecular Testing With Afirma GSC or Thyroseq v3
5
PubMed Central. Molecular Testing and Surgical Outcomes in Bethesda III and IV Thyroid Nodules: A Retrospective Cohort Study
6
PubMed. Molecular Profiling for Bethesda III to VI Nodules: Results of a Multicenter International Retrospective Study
7
PubMed Central. Management of Thyroid Nodules
8
PubMed Central. Prospective Study of Bethesda Categories III and IV Thyroid Nodules: Outcomes and Predictive Value of BRAF V600E Mutation
9
iScience. AI diagnosis of Bethesda category IV thyroid nodules
10
PubMed Central. A real-world experience of active surveillance in Bethesda IV thyroid nodules
11
PubMed Central. Assessment of thermal ablation for treating Bethesda IV thyroid nodules: a systematic review and meta-analysis
12
PubMed. Volume reduction rate of radiofrequency ablation in ≤ 2 cm Bethesda IV thyroid nodules
13
PubMed Central. Contrast-enhanced ultrasound combined with shear wave elastography in the diagnosis of C-TIRADS category 4 thyroid nodules
14
PubMed. Clinical Evaluation of an Artificial Intelligence-Based Decision Support System for the Diagnosis and American College of Radiology Thyroid Imaging Reporting and Data System Classification of Thyroid Nodules
15
PubMed Central. Risk Stratification of Thyroid Nodules: From Ultrasound Features to TIRADS