What Percent of Ground-Glass Nodules Are Cancerous?

Most ground-glass nodules found on CT scans are not cancerous. In large screening studies, the malignancy rate for pure ground-glass nodules (often called GGNs or GGOs) runs in the single digits, with data from the National Lung Screening Trial showing risks ranging from under 1% for the smallest nodules up to about 11% for those 20 mm or larger.¹ But those numbers shift dramatically depending on the nodule’s size, whether it has a solid component, whether it grows over time, and which population is being studied. The gap between a screening-detected GGN and one that ends up under a surgeon’s scalpel is enormous, and understanding where your nodule falls on that spectrum matters more than any single percentage.

Why the Reported Percentages Vary So Widely

If you search for the malignancy rate of ground-glass nodules, you will find numbers ranging from less than 1% to over 90%. Both ends of that range are technically correct, and neither tells you much on its own. The discrepancy comes down to who was being studied. In lung cancer screening programs, where thousands of healthy or at-risk people get low-dose CT scans, the vast majority of GGNs turn out to be harmless. Most are small, most never grow, and many vanish entirely on follow-up imaging. In those populations, the overall cancer rate among pure GGNs sits roughly in the range of 1% to 18%, depending on the screening protocol and how nodules were classified.¹

Contrast that with surgical series, where only nodules considered suspicious enough to warrant removal are included. In one study of resected ground-glass and part-solid nodules, the malignancy rate was over 93%.² That is not because GGNs are usually cancerous. It is because surgeons do not remove nodules unless they already have strong reasons to suspect cancer. The denominator has been pre-filtered, leaving mostly malignant cases. When your doctor quotes a number, the critical question is which population that number came from.

How Size Shapes Malignancy Risk

Size is one of the strongest single predictors of whether a ground-glass nodule is malignant. Data from the National Lung Screening Trial broke down GGO malignancy risk by diameter, and the gradient is steep. Nodules measuring 4 to 5 mm carried about a 0.4% risk of cancer. At 6 to 7 mm, the risk climbed to roughly 1.1%. Nodules between 8 and 14 mm had about a 3% chance, those from 15 to 19 mm about 5.2%, and those 20 mm or larger around 10.9%.¹ In practical terms, the majority of small GGNs are innocent bystanders. The updated Fleischner Society guidelines raised the threshold size for recommending follow-up imaging, reflecting the very low risk of tiny nodules.³

Research from a large Chinese screening program examined different diameter cutoffs for pure GGNs and found that an 8 mm threshold offered the best balance between catching invasive cancers and avoiding unnecessary workups. Raising the threshold from 5 mm to 8 mm improved specificity substantially while still detecting about three-quarters of invasive cases.⁴ Going above 10 mm actually started to reduce diagnostic accuracy for some cancer subtypes, suggesting there is a sweet spot around 7 to 8 mm where clinical suspicion should increase.

A meta-analysis confirmed that invasive adenocarcinoma presenting as a pure GGN tends to be significantly larger than preinvasive lesions.⁵ A separate study found that a larger diameter, the presence of a solid component, and a vascular convergence sign on CT were all independent risk factors for malignancy, with the combination achieving sensitivity above 90%.⁶

Many Ground-Glass Nodules Disappear on Their Own

One of the most reassuring facts about GGNs is that a large fraction are transient. They show up on one scan and are gone on the next, typically because they represented a minor infection, inflammation, or localized immune response rather than anything neoplastic. In one retrospective analysis, more than two-thirds of pure ground-glass opacities resolved during follow-up, and over half of mixed GGOs did too. The main culprits were pneumonia and eosinophilic lung infiltrates.⁷

A study examining 255 pure GGNs confirmed surgically or by follow-up found that 64 of them were transient lesions. The rest fell along a spectrum from atypical adenomatous hyperplasia (a precancerous condition) through adenocarcinoma in situ, minimally invasive adenocarcinoma, and fully invasive adenocarcinoma.⁸ This highlights an important point: even among persistent GGNs, many represent very early-stage or preinvasive disease that may never become a life-threatening cancer.

Pure Ground-Glass Versus Part-Solid Nodules

The distinction between a pure ground-glass nodule and a part-solid one matters enormously for risk. A pure GGN appears as a hazy patch on CT without any dense core. A part-solid nodule has both a ground-glass haze and a solid component within it. Part-solid nodules carry a substantially higher malignancy rate. One widely cited screening study reported that about 63% of part-solid nodules were malignant, compared with roughly 18% of pure ground-glass nodules.⁹

The development of a solid component within a previously pure GGN is one of the clearest red flags radiologists watch for. It suggests that a nodule is becoming more invasive, as the solid portion often corresponds to tumor cells replacing the normal air-filled lung tissue. When guidelines assign GGNs to different risk tiers, the presence or absence of a solid component consistently bumps a nodule into a higher management category. If your radiology report describes your nodule as “pure ground glass” with no solid part, that alone puts you in a lower-risk group than someone with a part-solid finding.

What Happens When a GGN Persists but Doesn’t Grow

Most persistent pure GGNs grow slowly or not at all. A study tracking pure GGO nodules detected by low-dose CT found that on a per-nodule basis, only about 10% showed growth, and those that did grow had a median volume doubling time of 769 days, or roughly two years.¹⁰ That is extremely slow compared with typical solid lung cancers, which often double in volume within a few months. In the same study, all 11 growing nodules that were surgically removed turned out to be primary lung cancers, confirming that growth is a reliable danger signal even when it happens slowly.

The MILD trial provided some of the longest follow-up data on ground-glass nodules under surveillance. The progression rate toward clinically relevant disease was extremely low, and researchers concluded that active surveillance is a safe strategy for most subsolid nodules.¹¹ Among the handful that were eventually resected after years of monitoring, seven of eight turned out to be stage I, and none of those patients died of lung cancer during a median follow-up of about four and a half years after surgery.¹² The takeaway is that even GGNs that eventually prove cancerous tend to be caught early enough that outcomes remain excellent, as long as surveillance continues.

CT Features That Raise or Lower Suspicion

Beyond size and solid components, radiologists look at a constellation of imaging features to gauge how worrisome a GGN looks. Malignant ground-glass nodules tend to display several telling signs on thin-section CT: tiny air-filled spaces within the nodule (vacuoles), air bronchus signs where a bronchiole passes through the lesion, vessels converging toward the nodule, spiculated or lobulated margins, and a tug on the nearby lung lining called pleural retraction.¹³ A meta-analysis confirmed that pleural traction and spiculation were both significantly more common in invasive adenocarcinoma than in preinvasive lesions, and that invasive tumors also showed higher average CT density values.⁵

Benign GGNs, by contrast, tend to look less organized. One study found they were often located in the upper lobes and had irregular shapes with coarse margins but generally well-defined borders.¹⁴ The difficulty is that none of these individual features is definitive on its own. There is considerable overlap between benign and malignant GGNs on imaging, which is why follow-up scans and clinical context matter as much as any single snapshot.

Who Gets GGNs and Who Is at Higher Risk

Ground-glass nodules are actually found more frequently in never-smokers than in smokers, which surprises many people who associate all lung nodules with smoking.¹⁵ However, the clinical significance of GGNs does differ by risk profile. Among people traditionally considered high-risk for lung cancer (older adults, heavy smokers), GGN-associated adenocarcinomas are more likely to be invasive. In lower-risk individuals, the same type of nodule more often represents a preinvasive lesion or minimally invasive adenocarcinoma.¹⁶ Both lesion size and smoking history have been identified as predictors of GGN growth over time.¹⁵

The implication is nuanced. Finding a GGN does not mean you need to be a smoker to worry, and it does not mean non-smokers are somehow protected from malignancy. But the probability that a given nodule will turn aggressive is shaped by the patient’s overall profile. Clinicians fold these risk factors into models like the Brock University or Mayo Clinic calculators to estimate the likelihood that a specific nodule is malignant, though these calculators have known limitations for subsolid nodules specifically.

Why PET Scans Often Don’t Help With Pure GGNs

Patients sometimes assume that a PET scan will settle the question of whether a nodule is cancerous. For solid lung nodules, PET scanning works well because cancers typically consume glucose at a high rate, making them light up on the scan. Pure GGNs are a different story. Because they are low-density and often represent very early or slowly growing disease, they tend to have low metabolic activity that is indistinguishable from inflammation. One study found no significant difference in PET uptake between malignant and inflammatory pure GGNs.¹⁷ This is a genuine diagnostic blind spot. If your doctor skips a PET scan for a small pure GGN, it is not because they are ignoring the nodule; it is because the scan would not provide useful information for that type of lesion.

Emerging Tools for Better Answers

The limitations of standard imaging have pushed researchers toward more sophisticated approaches. One active area is radiomics, which involves extracting hundreds of quantitative features from CT images that the human eye cannot easily perceive, then feeding them into prediction models. Studies combining radiomics features with clinical data have achieved diagnostic accuracy in the range of 70% to 86% for distinguishing benign from malignant GGNs, outperforming models built on either clinical data or imaging features alone.^{18 19} These tools are not yet part of routine clinical practice, but they represent a possible future where the wait-and-rescan cycle is shortened for at least some patients.

Another emerging approach is liquid biopsy, which looks for tumor-derived molecules circulating in the blood. A recent study identified a panel of four non-coding RNA markers in blood plasma that, when used together, achieved an area under the curve of 0.87 for distinguishing malignant from benign GGNs.²⁰ A simple blood draw that could flag high-risk nodules before they grow or change would be transformative, though these markers still need validation in larger, prospective studies before they could be used in clinical decisions.

When Surgery Is Needed, Outcomes Are Generally Excellent

For GGN-associated cancers that do need treatment, the prognosis is overwhelmingly favorable. These tumors are almost always adenocarcinomas, and they tend to be caught at stage I, before they have spread to lymph nodes or distant organs. A study comparing surgical approaches for small GGO-dominant lung adenocarcinomas (2 cm or smaller) found that both standard lobectomy and more limited resections produced equivalent overall survival.²¹ The ability to perform a smaller operation without compromising outcomes is significant, because it preserves more lung tissue and reduces surgical risk.

The indolent biology of most GGN-associated cancers is part of what makes surveillance safe for many patients. As the MILD trial demonstrated, even nodules monitored for years before resection were still overwhelmingly stage I at the time of surgery, and no patient died of lung cancer after treatment.¹² The window of opportunity to intervene rarely slams shut in months the way it can with aggressive solid tumors.

The Genetic Landscape of GGN Cancers

When GGN-associated lung adenocarcinomas are tested genetically, they frequently carry EGFR mutations, which are among the most common driver mutations in lung cancer and, importantly, are targetable with existing drugs. Studies have found EGFR mutation rates in GGO-associated tumors in the range of 50% to 57%.^{22 23} This is relevant not just for treatment planning if a nodule does prove to be cancer, but also for understanding why these tumors behave differently from the stereotypical aggressive lung cancers associated with heavy smoking. EGFR-driven cancers tend to arise in never-smokers or light smokers, grow slowly, and respond well to targeted therapies.

Living With an Indeterminate Nodule

The medical reality of ground-glass nodules is largely reassuring: most are benign, the cancerous ones are usually caught early, and outcomes after treatment are excellent. The psychological reality is harder. Research on patients undergoing surveillance for screen-detected nodules found that learning about a lung nodule frequently triggered distress, anxiety, and fear about cancer. People described feeling shocked, scared, and struggling to cope with the uncertainty of waiting months for the next scan.²⁴

That anxiety is understandable, but it is worth calibrating against the actual risk. If you have a small pure ground-glass nodule with no solid component and no worrisome CT features, the chance it is cancer is in the low single digits, and the chance it is an aggressive cancer is even lower. The surveillance interval your doctor recommends, whether six months or a year, is not a period of danger; it is a deliberate strategy that takes advantage of the slow biology of these lesions. Growing nodules will reveal themselves over time, and the evidence consistently shows that waiting does not turn a curable cancer into an incurable one for GGN-type disease. If the next scan looks the same, that itself is meaningful information that should bring genuine reassurance.