Does Cortical Thickening of a Lymph Node Mean Cancer?

Cortical thickening of a lymph node does not, by itself, mean cancer. It is one of several features radiologists evaluate when assessing whether a node looks suspicious, but the finding is common in entirely benign conditions ranging from recent vaccination to routine infections. The specifics matter: how thick the cortex is, whether the thickening is uniform or lopsided, whether the node’s central fatty hilum is still visible, and what else is going on clinically. Most people whose imaging report mentions cortical thickening will turn out not to have cancer, but the finding does warrant context and sometimes follow-up.

What “Cortical Thickening” Actually Describes

A lymph node, when viewed on ultrasound, has a characteristic kidney-bean shape. The outer rim is the cortex, the tissue where immune cells congregate and respond to infections or other threats. The center is a bright, fatty area called the hilum, where blood vessels enter and exit. In a healthy node, the cortex is a thin, even rind around a clearly visible hilum.

When radiologists say the cortex is “thickened,” they mean that outer rim has grown wider than expected. This happens whenever the node is working harder, whether because it’s fighting off a nearby infection, reacting to a vaccine, responding to an inflammatory skin condition, or harboring cancer cells that have traveled from a tumor. On imaging alone, the node doesn’t announce why it’s thickened.

Normal cortical width varies by body region. In one study of healthy volunteers, the cortex in axillary nodes (the armpit) measured up to 4 mm in 95% of normal subjects, while inguinal nodes (the groin) stayed at 2.5 mm or less in the same proportion.¹ That regional difference matters: a 3.5 mm cortex in an armpit node may be perfectly normal, while the same measurement in a groin node could be flagged as abnormal.

How Thick Is Too Thick

There is no single agreed-upon number at which cortical thickening becomes “suspicious.” Different research groups have tested different cutoffs, and the trade-off is always the same: set the threshold low and you’ll catch more cancers but also flag more benign nodes; set it high and you’ll miss fewer benign nodes but also miss some cancers.

In breast cancer staging, which is where most of this research has been done, a commonly studied cutoff is 3 mm. One study found that using cortical thickness above 3 mm as the criterion for a suspicious axillary node yielded about 67% sensitivity and 75% specificity, and that cortical thickness was the single strongest independent predictor of nodal metastasis on multivariate analysis.² A different study used a lower cutoff of 2.5 mm and achieved higher sensitivity at 85%, with 78% specificity.³

A more recent study of over 500 axillary nodes placed the optimal cutoff higher, at 5 mm, and found that cortical thickness alone had the highest discriminative ability among all ultrasound features they measured, with 90% sensitivity and 71% specificity.⁴ The spread in cutoffs across studies (2.5 mm, 3 mm, 5 mm) reflects genuine uncertainty about the best threshold and the reality that institutions differ in how aggressively they want to biopsy.

The practical upshot: if your imaging report says cortical thickening of 3 or 4 mm in an axillary node, that falls in a gray zone. It could be completely normal reactivity or it could be early metastatic involvement. Doctors don’t usually make decisions based on thickness alone.

The Pattern Matters More Than the Number

Radiologists look at how the cortex is thickened, not just how much. A smooth, even rim of thickening all the way around the node looks different from a lopsided bulge on one side. That distinction carries real diagnostic weight.

In metastatic disease, cancer cells typically enter a lymph node through one of the lymphatic channels on its periphery, which means the tumor initially deposits on one side of the cortex. This produces asymmetric or eccentric thickening, a focal bulge rather than uniform expansion. One study of over 200 metastatic axillary nodes found that asymmetric cortical thickening on the side closer to the primary breast tumor was an independent predictor of metastasis, along with the ratio of cortex to hilum thickness and abnormal blood-vessel patterns within the node.⁵

Diffuse, smooth cortical thickening, by contrast, is the pattern most commonly associated with reactive (benign) causes. The immune system responds broadly, expanding the cortex evenly rather than bulging at one spot. That said, diffuse thickening is not a clean pass: the same study found diffuse thickening in more than half of the metastatic nodes examined, making it the most common pattern even in cancer. The distinction is probabilistic, not absolute.

Loss of the fatty hilum is another red flag. When the cortex expands enough to crowd out or replace the central hilum entirely, the node takes on a round rather than oval shape and loses that characteristic bright center on ultrasound. One study found the positive predictive value for malignancy was substantially higher in nodes lacking a normal hilum (61%) compared to those with a preserved hilum (23%), and that cortical thickness above 5 mm combined with hilum absence had the highest predictive value for metastatic disease.⁶

Benign Causes That Mimic Suspicious Nodes

The immune system doesn’t thicken a node’s cortex only in response to cancer. Several common, harmless situations produce the same imaging appearance, and recognizing them can spare you unnecessary anxiety and biopsies.

COVID-19 mRNA vaccination became a well-known cause of cortical thickening starting in 2021. The nodes in the armpit on the same side as the injection swell as the immune system responds to the vaccine. Ultrasound studies showed that this thickening, along with increased blood flow within the node, could persist for months. One study documented that the changes approached normal by about 24 to 28 weeks after vaccination.⁷ Another study modeling the rate of resolution estimated a half-life of about 77 days for the decrease in cortical thickness, and noted that diffuse, smooth cortical thickening above 3 mm was the most common appearance.⁸ If you had an mRNA vaccination within the past several months and your imaging shows thickened axillary nodes on the same side, that is very likely the explanation.

Routine infections are an even more common cause. A cut, bug bite, or skin infection on your arm can make the armpit nodes enlarge and thicken their cortices. Upper respiratory infections can do the same to cervical (neck) nodes. These reactive nodes tend to be oval, have preserved hila, and show smooth cortical thickening, but they can look worrisome on a scan, especially if you didn’t mention the infection to your doctor before the imaging was done.

A less well-known mimic is dermatopathic lymphadenitis, a benign reaction pattern where lymph nodes enlarge in response to chronic skin conditions like eczema, psoriasis, or other dermatoses. The node’s paracortical region expands with immune cells and even melanin-containing macrophages from the skin, producing nodes that can look suspicious enough to be confused with metastatic disease or even lymphoma.⁹ The takeaway is that skin problems you might not even think to mention can affect how your lymph nodes appear on imaging.

Age Changes the Baseline

What looks normal in a lymph node shifts with age. A study comparing ultrasound features of axillary nodes in younger and older healthy volunteers found that younger people had significantly thicker cortices, higher Bedi classifications (a scoring system for node appearance), and more blood flow visible on Doppler imaging.¹⁰ In other words, a 25-year-old’s normal nodes may look “suspicious” by criteria developed in studies of older breast cancer patients. This is one reason radiologists factor in the full clinical picture rather than applying a single thickness cutoff to everyone.

Conversely, older adults tend to have thinner cortices and more prominent fatty hila, making any cortical thickening in that age group somewhat more noteworthy. The immune system becomes less active with age, and the lymph node architecture gradually shifts toward more fat and less functional tissue. When an older person’s node does thicken, there’s less likely to be a vigorous benign immune reaction behind it and relatively more reason to investigate.

How Most Incidental Findings Turn Out

Many people first learn about cortical thickening when an imaging scan done for another reason catches a lymph node that looks a bit off. These incidental findings are common and overwhelmingly benign. One evaluation of incidentally discovered lymph nodes found that about 8.5% turned out to be malignant, while 91.5% were benign.¹¹ That study also made a sobering observation about the limits of imaging features: no single feature reliably separated the benign nodes from the malignant ones, and nearly all nodes had at least one feature that two independent radiologists considered suspicious. In other words, finding “something suspicious” on imaging is the norm, not the exception, and the vast majority of those findings are nothing.

The study also found that in many of the malignant cases, the cancer diagnosis ultimately came not from the node itself getting bigger but from the later appearance of a non-nodal tumor site. Sometimes a node that initially looks indeterminate just needs time and repeat imaging to reveal its true nature.

When Biopsy Becomes the Next Step

If a node looks suspicious enough on imaging, doctors often move to tissue sampling rather than simply watching and waiting. The two main options are fine needle aspiration (FNA) and core needle biopsy, both performed under ultrasound guidance.

A comparison of the two methods in patients with suspected axillary metastasis from breast cancer found that core biopsy detected cancer in about 87% of cases where metastasis was present, compared to roughly 79% for FNA. The results were discordant in about 20% of patients, with core biopsy catching some cases FNA missed and vice versa.¹² Many centers now prefer core biopsy for axillary nodes because the larger tissue sample allows not just a cancer-or-not answer but also receptor testing that guides treatment decisions.

Not every thickened node gets biopsied, though. Newer diagnostic strategies are trying to reduce unnecessary procedures by adding layers of assessment. One approach tested a stepwise system for cervical (neck) nodes: nodes scored as indeterminate on standard ultrasound underwent contrast-enhanced ultrasound, and their classification was then upgraded toward biopsy or downgraded toward follow-up based on how the node enhanced.¹³ The goal of systems like this is to spare patients with clearly benign-appearing nodes from invasive procedures while still catching the ones that need tissue diagnosis.

Beyond Ultrasound

Ultrasound is usually the first tool used to evaluate a lymph node because it’s cheap, fast, and radiation-free. But when the clinical question is whether cancer has spread to lymph nodes, other imaging methods come into play, and each has different strengths.

A systematic review comparing CT, MRI, PET-CT, and ultrasound for detecting cervical lymph node metastases in head and neck cancer found that PET-CT had the strongest overall diagnostic performance at the patient level, with about 75% sensitivity and 84% specificity. MRI had the highest sensitivity on a per-node basis at roughly 77%, while PET alone had the highest specificity at nearly 97%.¹⁴ No single modality is perfect, and in practice doctors often combine information from multiple types of imaging along with clinical history.

For people who have had a PET-CT that shows equivocal lymph node findings, the waiting period can be particularly stressful. A study of patients with indeterminate lesions on PSMA PET-CT (used in prostate cancer staging) found that follow-up assessment changed clinical management in nearly half the patients, with many initially equivocal nodes eventually reclassified as benign.¹⁵ That high reclassification rate reinforces the point that “indeterminate” on an initial scan often resolves to “benign” with a bit of time.

Artificial Intelligence and Where Diagnosis Is Heading

Researchers are actively training machine learning models to read lymph node ultrasound images, hoping to outperform the inherent subjectivity of human interpretation. The results so far are promising. One model that combined conventional ultrasound features with contrast-enhanced ultrasound data achieved about 93% accuracy in predicting axillary node metastasis in breast cancer, and ranked diffuse cortical thickening above 3 mm and eccentric cortical thickening above 3 mm among the most important features driving its predictions.¹⁶

Another study built a machine learning model for cervical lymph nodes using a broad set of ultrasound features and found it significantly outperformed an experienced human ultrasound specialist, with an area under the curve of 0.987 compared to the physician’s 0.904. The model identified cortical thickness as one of many features contributing to its decisions, alongside shape, blood flow patterns, and internal echogenicity.¹⁷ These tools aren’t replacing radiologists yet, but they highlight an important principle already at work in human decision-making: cortical thickness is just one input in a much larger equation, and the best diagnostic performance comes from weighing many features together rather than relying on any single measurement.

What to Tell Your Doctor

If your imaging report mentions cortical thickening, a few pieces of clinical context can help your doctor interpret the finding accurately and potentially save you from unnecessary procedures. Recent vaccinations, especially mRNA COVID vaccines in the arm on the same side as the enlarged node, are worth mentioning even if the shot was months ago, since those changes can linger for half a year or more.⁷ Recent infections, cuts, or skin conditions near the area drained by the node are also relevant. So is your age: a mildly thickened node in a 30-year-old with no cancer history carries a very different probability than the same finding in a 65-year-old being staged for a known tumor.

The report’s language can also give you clues about how concerned the radiologist actually is. Phrases like “mildly prominent” or “likely reactive” suggest the radiologist thinks the finding is benign and may recommend short-term follow-up at most. Phrases like “eccentric cortical thickening,” “loss of fatty hilum,” or “suspicious morphology” carry more weight and usually lead to a recommendation for biopsy or further imaging. If the report recommends follow-up in three to six months rather than immediate biopsy, that generally means the finding is indeterminate but not alarming enough to warrant tissue sampling right away.