A fatty hilum on imaging makes a lymph node cancerous in only a small minority of cases — roughly 1% to 5% in most biopsy series of sonographically benign nodes — because the hilum is the node's normal exit door for fat and lymphatic fluid, and cancer does not build one. It destroys one. When the hilum is preserved, thin, and echogenic (bright on ultrasound), the odds of malignancy drop sharply. Benign fatty-hilum nodes are common, especially in the neck, armpit, and groin. The rare exceptions are nodes with a normal hilum but abnormal cortex, and those are where missed cancers hide. This article explains the mechanism, the numbers, and the exceptions.
What a Fatty Hilum Actually Is
The hilum is not a fat deposit that grows on a lymph node like moss on a rock. It is the node's anatomical exhaust port — the concave side where the efferent lymphatic vessel leaves and where blood vessels enter and exit.
Fat enters through that same door as the node involutes with age. Medullary fat replaces lymphoid tissue from the hilum outward, which is why older adults have more fatty-replaced nodes than children. On ultrasound, fat is echogenic because fat-water interfaces reflect sound, so a normal hilum appears as a bright, narrow, wedge-shaped or linear band, 1 to 3 mm thick, continuous with the surrounding perinodal fat.
There are three practical imaging signatures a radiologist checks. On ultrasound, a preserved echogenic hilum with a thin, smooth, hypoechoic cortex under 3 mm. On CT, a low-attenuation fatty center (roughly minus 80 to minus 100 Hounsfield units) with a smooth cortical rim. On MRI, T1 hyperintense fat in the hilum mirroring subcutaneous fat. In plain terms: a fatty hilum is the node's normal front door, visible because fat looks different from the surrounding lymph tissue.
Why Cancer Tends to Erase the Hilum
Suspicious lymph nodes do not usually lose their fat because of something benign; they lose it because tumor cells and inflammation both remodel the node from the cortex inward.
Here is the counterintuitive part. Malignant cells typically enter a node through the afferent lymphatics into the subcapsular sinus, on the opposite side from the hilum. The tumor grows as an eccentric cortical mass, or as diffuse effacement, and it compresses and eventually obliterates the medullary fat from the inside out. So a preserved hilum is essentially a report that the tumor has not yet reached the center of the node — a real, but temporary, structural tell.
That gives us a mechanistic reason to trust the sign. It also gives us our first caveat: a tumor that starts near the hilum, or a metastasized node with only a small peripheral deposit, can spare the fat while already being malignant. This is the "normal-appearing hilum" edge case and it is rare, but it exists.
Inflammation works differently. Reactive follicular hyperplasia enlarges the cortex and preserves the hilum; the node grows symmetrically around its exit door. Infection can thicken the cortex too. So both reactive nodes and early tumors can enlarge — the difference is geometry, not size alone.
The Numbers: How Much a Fatty Hilum Lowers Cancer Risk
The strongest data comes from ultrasound-guided fine-needle aspiration series. In published cohorts of cervical nodes classified as benign by sonographic criteria (short axis under 8 mm, thin cortex, preserved echogenic hilum, and hilar flow), malignancy is found in a small percentage — often quoted between 1% and 5%, and lower when all four criteria are met.
| Finding on Ultrasound | Typical Malignancy Rate in Biopsy Series | What It Means |
|---|---|---|
| Preserved fatty hilum alone | 1% to 5% | Low risk but not zero |
| Preserved hilum + short axis under 8 mm + thin cortex | Around 1% or less | Very low risk; often observed |
| Absent hilum, round shape, short axis over 10 mm | 50% to 90% | High risk; biopsy indicated |
| Preserved hilum + abnormal cortex (eccentric focal thickening, lost hilar flow) | Elevated, variable | Biopsy recommended despite benign hilum |
The last row is the one that matters most, and it is where most thin explainers stop. A fatty hilum does not override a bad cortex. The 2014 American College of Radiology TI-RADS white paper and ultrasound criteria for cervical nodes both treat hilar loss as one feature among several, not a single deciding test.
A second number worth knowing: the short-axis diameter threshold of 8 mm to 10 mm is where observational data separates "normal-sized" from "likely abnormal" nodes in most body regions. In the neck, 8 mm is the common cutoff; in the groin, normal nodes can reach 15 mm short axis without concern because the benign size ceiling is higher there. Same rule, different numbers by body region.
Where the Rule Breaks: Edge Cases That Matter
A fatty hilum is reassuring, but the exceptions cluster in specific places and populations. Here are the ones a careful radiologist watches.
- Size: a node with a preserved hilum but short axis over 10 mm in the neck, or over 15 mm in the groin, still deserves attention even if the hilum looks normal.
- Cortex: eccentric cortical thickening of more than 3 mm, lobulation, or a cortex that becomes hypoechoic on one side only.
- Dopper flow: hilar vascularity is normal; displaced or chaotic cortical flow is not.
- Location: supraclavicular nodes above 5 mm short axis are suspicious almost regardless of hilum, because that region drains the chest and abdomen.
- Clinical setting: a known primary cancer elsewhere shifts the pretest probability dramatically — a preserved hilum in a patient with a known head and neck cancer carrier still warrants biopsy.
- Calcification and cystic change: microcalcified or necrotic nodes can be metastatic even with visible hilar fat, particularly in papillary thyroid carcinoma and in some lymphomas.
In plain terms: the hilum is one vote, not a veto. Size, cortex, flow, location, and the patient's history are the other votes, and the radiologist is tallying them all.
Why Size Alone Fails and the Hilum Beats It
Many patients believe node size is the main signal. It is not, and the correction is important.
A reactive node in the neck can reach 12 mm short axis and be entirely benign. A 6 mm node with a lost hilum and a thick cortex can be metastatic. Size is a weak predictor of malignancy in isolation because both inflammation and tumor enlarge nodes; the shape and internal architecture are what separate them. A skinny node with a fatty center is reassuring; a plump round node with no fat center is not, even if it is smaller.
There is a second misconception worth correcting. Patients often assume a fatty hilum means the node is "just a fat lump" or a lipoma. That is not accurate. A fatty hilum node is a normal lymph node with fatty replacement of its medulla, not a fat tumor. Lipomas and other fat-containing soft tissue masses are a different entity and are typically fully homogeneous fat without any lymphoid cortex at all.
A third misconception: "If it has fat in it, it cannot be cancer." Fat necrosis, a benign process, can mimic malignancy and produce fat-containing masses with irregular margins after trauma or surgery — but that is a lesion in fat, not a node with a hilum.
What to Actually Do When You See the Report
The practical steps depend on where the node is, how big it is, and what else the report shows.
Ask for the full ultrasound description, not just the conclusion. You want the short-axis measurement, the cortical thickness, whether the hilum is present or lost, the vascular pattern, and any abnormality in adjacent tissue. The phrase "normal fatty hilum" by itself does not tell you the other variables.
If the report says preserved hilum, short axis under 8 mm, thin uniform cortex, no suspicious features, and no known cancer: observation is reasonable, and the short-term follow-up ultrasound at about 3 to 6 months is standard in many guidelines rather than immediate biopsy. A growing node on that interval imaging changes the plan.
If the report says preserved hilum but any of the following are present, ask your clinician whether biopsy is warranted rather than waiting: short axis over 10 mm in the neck or over 15 mm in the groin, eccentric cortical thickening, displaced cortical flow, a supraclavicular location, or a known primary tumor elsewhere. In those settings, a fine-needle aspiration or core biopsy gives a tissue answer in under a week in most centers.
If the node has no visible hilum, is round, or shows cortical calcification, expect a biopsy recommendation. A 30 to 60 minute outpatient fine-needle aspiration with results in 2 to 7 days is the typical path.
For palpable neck nodes that persist beyond 4 to 6 weeks without a clear infectious cause, imaging is indicated regardless of what the first ultrasound said, because a small fraction of normal-appearing nodes later turn out to harbor malignancy.
The Limits of Reassurance
No imaging sign replaces tissue. The fatty hilum is one of the most useful benign features in lymph node imaging, but it is a Bayesian signal, not a diagnosis. In a young person with a soft, mobile, 5 mm neck node and a preserved hilum, the post-test probability of malignancy is very low. In a 60-year-old with a known lung or head and neck cancer and a 12 mm supraclavicular node with a maintained hilum, it is not low at all — the same appearance, different answer, because the pretest probability moved.
That is the honest limit of the science. The fat in the middle of a lymph node is normal, most of the time harmless, and usually a good sign. The reason a doctor still wants the measurements, the flow pattern, the location, and the clinical story is that the node's geometry is a single clue, and medicine reads clues in groups. Per the American College of Radiology ultrasound criteria and the 2014 TI-RADS framework, the interpretation is contextual, not a single-box checklist. If you are uncertain after reading your report, the right next move is a conversation with the radiologist who interpreted it, not a second opinion from a search engine.