Roughly 5% of thyroid fine-needle aspiration (FNA) biopsies return a clearly malignant result. In a study of nearly 4,000 patients who received diagnostic FNA results, 5.3% were malignant, 87.5% were benign, and about 7% landed in an indeterminate gray zone that required further workup.1PubMed Central. The incidence of thyroid cancer by FNA varies by age and gender That one-in-twenty headline number, though, obscures a more complicated picture shaped by your age, sex, ultrasound appearance of the nodule, and what the pathologist sees under the microscope.
How Biopsy Results Are Categorized
Thyroid biopsy results are not simply “cancer” or “no cancer.” Pathologists use a six-tier grading system called the Bethesda System for Reporting Thyroid Cytopathology, which assigns each sample to a category that carries its own estimated risk of malignancy. Understanding which category your biopsy falls into matters far more than a blanket percentage.
At the extremes, the system works well. Category II (benign) carries a malignancy risk below 2%, and Category VI (malignant) is confirmed cancer essentially 100% of the time. One institutional study found malignancy rates of 1.5% for benign readings, 86.5% for “suspicious for malignancy” (Category V), and 100% for Category VI.2Cirugía Española (English Edition). Efficiency of the Bethesda System for Thyroid Cytopathology For most people whose biopsy returns benign, the chance of a missed cancer is very low.
Category I covers samples that are “non-diagnostic,” meaning the needle did not capture enough usable cells. This happens in roughly 5 to 15% of biopsies, and the standard recommendation is to repeat the procedure. Among patients with non-diagnostic results who were followed over time, about 3% were eventually found to have thyroid cancer. The strongest predictor of malignancy in those cases was the presence of calcifications inside the nodule, while larger nodules were actually less likely to be cancerous.3PubMed Central. Risk of Malignancy in Thyroid Nodules with Non-Diagnostic Fine-Needle Aspiration: A Retrospective Cohort Study Patients who had repeated non-diagnostic results and never went to surgery were not diagnosed with cancer during follow-up in that study, which is reassuring if you are stuck in a cycle of inconclusive biopsies.
The Indeterminate Gray Zone
The most anxiety-producing territory is Categories III and IV, the indeterminate results. Category III, labeled “atypia of undetermined significance” or “follicular lesion of undetermined significance” (AUS/FLUS), was originally designed to carry a 5 to 15% cancer risk. But real-world data has been less tidy.
A pooled analysis of AUS/FLUS nodules found a confirmed malignancy rate of about 27% across all such nodules, and among those that went directly to surgery, the rate climbed to roughly 38%.4PubMed Central. Malignancy rate in thyroid nodules classified as Bethesda category III (AUS/FLUS) – Section: Abstract Other centers have reported a wide spread. One tertiary care center found the malignancy rate in AUS/FLUS nodules to be between 13 and 25%, with every single cancer turning out to be a variant of papillary thyroid carcinoma.5PubMed Central. Rate of malignancy for thyroid nodules with AUS/FLUS cytopathology in a tertiary care center – a retrospective cohort study A review of recent literature suggested that some centers see rates as high as 38 to 55% for this category, well above what was originally expected.6PubMed Central. Atypia of undetermined significance/follicular lesions of undetermined significance: What radiologists need to know
Why the enormous range? Part of it comes down to how liberally different institutions use the AUS/FLUS label. A center that uses the category sparingly, reserving it for truly borderline cases, will naturally accumulate a higher proportion of cancers in that bin. A center that uses the label more broadly as a catch-all for slightly unusual cells will have a lower malignancy rate. The bottom line for patients is that an indeterminate result does not mean cancer, but it does mean the biopsy alone cannot give a clear answer. The next steps typically involve either a repeat biopsy, molecular testing, or surgery.
Age, Sex, and Who Faces Higher Odds
Your demographics shift the math. In the same large study mentioned above, patients younger than 45 had malignant biopsy rates roughly twice as high as older patients: about 8% versus 4%.1PubMed Central. The incidence of thyroid cancer by FNA varies by age and gender That feels counterintuitive, since cancer in general is thought of as a disease of aging. But thyroid cancer is unusual: it peaks in middle-aged adults, and younger people who develop nodules suspicious enough to biopsy are more likely to have an actual malignancy.
Men had a higher malignancy rate on biopsy than women (about 7% versus 5%), and also had more indeterminate results (about 10% versus 6%). Women develop thyroid nodules far more often than men, so the total number of thyroid cancers is higher in women. But on a per-biopsy basis, a man who reaches the point of having a needle stuck in his thyroid is somewhat more likely to hear bad news. Pediatric thyroid nodules carry an even higher per-biopsy cancer risk than adult ones, which is why some guidelines recommend a more cautious approach to evaluating thyroid lumps in children.7PubMed Central. Incidence and malignancy rates of indeterminate pediatric thyroid nodules
What Happens Before the Biopsy Even Occurs
Most thyroid nodules are never biopsied. Ultrasound scoring systems, the most widely used being ACR TI-RADS (Thyroid Imaging Reporting and Data System), evaluate features like shape, echogenicity, margins, and the presence of calcifications to assign a suspicion level. Only nodules above a certain suspicion threshold, and above a certain size for their suspicion level, are recommended for biopsy. This filtering step has a big effect on the final malignancy percentage, because higher-suspicion nodules are more likely to be biopsied and more likely to be cancer.
A Canadian study of over 1,000 consecutive FNA biopsies stratified by TI-RADS level showed how dramatically risk varies. Nodules classified as “benign” on imaging (TR1 and TR2) had a 0% cancer rate. “Mildly suspicious” nodules (TR3) had a cancer rate of 0.3%. “Moderately suspicious” nodules (TR4) had a rate of about 5%, and “highly suspicious” nodules (TR5) had a cancer rate above 40%.8PubMed Central. Cancer Risk in Thyroid Nodules: An Analysis of Over 1000 Consecutive FNA Biopsies Performed in a Single Canadian Institution TI-RADS 4 or higher detected malignant nodules with about 92% sensitivity, though specificity was only about 53%, meaning it correctly flagged most cancers but also flagged many benign nodules.9PubMed. Prediction of thyroid nodule malignancy using thyroid imaging reporting and data system (TIRADS) and nodule size
Different countries use slightly different ultrasound classification schemes (EU-TIRADS, Kwak-TIRADS, KTA/KSThR), and their performance varies. One comparative study found that pattern-based systems tend to be more sensitive (catching more cancers) while point-scoring systems tend to be more specific (flagging fewer benign nodules unnecessarily), but all of them lead to a substantial number of biopsies that turn out to be benign.10PubMed. Pattern-based vs. score-based guidelines using ultrasound features have different strengths in risk stratification of thyroid nodules That trade-off between catching cancers and avoiding unnecessary procedures is a running theme in thyroid medicine.
Molecular Testing for Indeterminate Nodules
When a biopsy comes back indeterminate, you used to face a binary choice: repeat the biopsy or go to surgery to find out whether it was cancer. Molecular testing has opened a third option. These tests analyze the genetic material in the biopsy sample, looking for mutations and gene expression patterns associated with thyroid cancer. The goal is not to replace the biopsy but to help “rule out” cancer in the indeterminate cases, potentially sparing you an operation.
A randomized clinical trial comparing two molecular approaches found that both had very high sensitivity, meaning they rarely missed a cancer. The RNA-based test caught 100% of malignancies, and a combined DNA-RNA test caught about 97%. Specificity was more moderate, around 80 to 85%, meaning some benign nodules still tested positive. About half of patients with indeterminate nodules had negative molecular test results and were able to avoid diagnostic surgery.11JAMA Oncology. Effectiveness of Molecular Testing Techniques for Diagnosis of Indeterminate Thyroid Nodules: A Randomized Clinical Trial
A meta-analysis pooling data on two widely used second-generation molecular tests found that both achieved sensitivities above 95% and negative predictive values in the low-to-mid 90s, meaning a negative result is quite reliable for ruling out cancer.12PubMed. Diagnostic performance of the second-generation molecular tests in the assessment of indeterminate thyroid nodules: A systematic review and meta-analysis A health technology assessment concluded that molecular testing reduces unnecessary surgeries compared to the traditional approach of diagnostic lobectomy, with a slight improvement in quality-adjusted life years, though it does increase costs.13PubMed Central. Molecular Testing for Thyroid Nodules of Indeterminate Cytology: A Health Technology Assessment
Molecular testing is not perfect. Specificity hovers around 50 to 85% depending on the test and the study, so a positive result does not guarantee cancer. It still tends to push patients toward surgery, though with more information in hand. These tests are most valuable as “rule-out” tools: a negative result lets many patients safely avoid the operating room.
Core Needle Biopsy as an Alternative
Standard thyroid biopsy uses a fine needle that aspirates loose cells. Core needle biopsy (CNB) uses a slightly thicker needle that retrieves an intact sliver of tissue, preserving the architecture of the sample. A systematic review and meta-analysis found that CNB yielded far fewer non-diagnostic results compared to FNA, with a pooled risk of a non-diagnostic sample about 73% lower for core needle biopsy.14PubMed Central. Comparison of diagnostic yield of core-needle and fine-needle aspiration biopsies of thyroid lesions: Systematic review and meta-analysis
A large comparison study found that CNB had a non-diagnostic rate of about 3% versus 11% for FNA, and fewer indeterminate AUS/FLUS results as well (about 1% versus 6%). CNB also classified more nodules into the benign category (about 61% versus 50%) and was better at detecting neoplastic disease overall. In terms of distinguishing cancer from non-cancer, though, the two approaches were not significantly different.15Endocrinology and Metabolism. Diagnostic Performance of Thyroid Core Needle Biopsy Using the Revised Reporting System: Comparison with Fine Needle Aspiration Cytology FNA remains the global standard for initial biopsy, but CNB is increasingly used in some countries, particularly South Korea, as a complement or follow-up when FNA is inconclusive.
How Much the Pathologist Matters
One uncomfortable reality in thyroid biopsy diagnosis is that pathologists do not always agree with each other. A prospective study found that when local pathologists and central expert cytopathologists reviewed the same slides, they agreed on the Bethesda category only 64% of the time. Even when the same pathologist re-read their own slides, agreement was only about 75%.16PubMed. A prospective assessment defining the limitations of thyroid nodule pathologic evaluation Central cytopathologists with more subspecialty expertise made fewer indeterminate calls (about 41%) compared to local pathologists (about 55%), suggesting that experience helps resolve borderline cases.
Another study found that the six-category Bethesda system had considerable overlap between certain adjacent categories, particularly between AUS/FLUS and “follicular neoplasm” and between “suspicious for malignancy” and “malignant.” A simplified four-category scheme improved agreement between different readers.17PubMed. A simplified Bethesda System for reporting thyroid cytopathology using only four categories improves intra- and inter-observer diagnostic agreement and provides non-overlapping estimates of malignancy risks This variability is worth knowing about if you receive an indeterminate result: a second opinion from a subspecialty cytopathologist can sometimes change the category, and with it, your recommended treatment path.
Most Thyroid Cancers Found on Biopsy Are Papillary
When thyroid biopsies do find cancer, the overwhelming majority are papillary thyroid carcinoma. In one large surgical series, about 75% of confirmed malignancies were papillary thyroid cancer.18PubMed. The accuracy of fine-needle aspiration biopsy and frozen section in patients with thyroid cancer Papillary thyroid cancer is generally slow-growing and has an excellent prognosis, with five-year survival rates above 98% for localized disease. Other types, including follicular, medullary, and anaplastic thyroid cancer, are less common and carry varying prognoses, with anaplastic being the most aggressive and rarest.
This dominance of papillary thyroid cancer has driven a reclassification that quietly lowered the cancer rate. In 2016, a subset of what used to be called “follicular variant of papillary thyroid carcinoma” was renamed to “noninvasive follicular thyroid neoplasm with papillary-like nuclear features” (NIFTP), essentially reclassifying it from cancer to a benign neoplasm. A meta-analysis estimated this reclassification affected about 31,000 patients annually worldwide, with the largest impact in North America and Europe.19PubMed. The Incidence of Noninvasive Follicular Thyroid Neoplasm with Papillary-Like Nuclear Features: A Meta-Analysis Assessing Worldwide Impact of the Reclassification In the United States, the incidence of follicular variant papillary thyroid cancer dropped sharply after 2015 as a result, though the observed increases in the newly named NIFTP category accounted for only about 20% of the decline, suggesting reporting practices shifted in complex ways.20The Journal of Clinical Endocrinology & Metabolism. Influence of Nomenclature Changes on Trends in Papillary Thyroid Cancer Incidence in the United States, 2000 to 2017
The Overdiagnosis Problem
The percentage of thyroid biopsies that are cancerous does not exist in a vacuum. How many biopsies are performed, and on whom, shapes the denominator. Over the past few decades, thyroid cancer diagnoses have risen dramatically in many countries, but death rates have stayed essentially flat. This mismatch points to widespread overdiagnosis: the detection of cancers that would never have caused symptoms or harm during a person’s lifetime.
A 2025 modeling study estimated that between 1991 and 2019, 72 to 94% of papillary thyroid cancer cases diagnosed in the United States were overdiagnosed, meaning they did not affect population mortality. Women were overdiagnosed at higher rates than men (75 to 95% versus 63 to 90%).21PubMed Central. Overdiagnosis of Papillary Thyroid Cancer A population-based temporal trend study identified three mechanisms driving these extra diagnoses: opportunistic screening during routine physical exams, diagnostic cascades where vague complaints trigger multiple tests, and incidental findings on imaging performed for unrelated reasons like CT scans of the chest or neck ultrasounds for carotid artery disease.22PLoS ONE. Overdiagnosis and overtreatment of thyroid cancer: A population-based temporal trend study Autopsy studies have long shown that a large reservoir of tiny papillary thyroid cancers exists in people who died of something entirely unrelated.
A recent review in Nature Reviews Endocrinology attributed much of the incidence increase to inappropriate use of imaging, especially thyroid and neck ultrasound.23PubMed Central. Unravelling the rise in thyroid cancer incidence and addressing overdiagnosis The practical consequence is that the 5% malignancy rate on biopsy includes many tiny, indolent papillary cancers that might never have been found in an era with less imaging, and that might never have caused any problem.
Active Surveillance for Tiny Papillary Cancers
Growing awareness of overdiagnosis has fueled interest in active surveillance as an alternative to immediate surgery for small, low-risk papillary thyroid microcarcinomas (tumors 1 cm or smaller). The concept is simple: rather than operating right away, monitor the nodule with periodic ultrasounds and only intervene if it grows or shows signs of spread.
The longest-running evidence comes from a Japanese study with 30 years of follow-up covering more than 3,200 patients in an active surveillance group. Over that period, only about 4% had tumor growth of 3 mm or more, with 10-year and 20-year enlargement rates of roughly 5% and 7% respectively. New lymph node metastases occurred in less than 1% of patients. No patients in the study died of thyroid cancer, whether they were in the surveillance group or the immediate surgery group.24PubMed Central. Long-Term Outcomes of Active Surveillance and Immediate Surgery for Adult Patients with Low-Risk Papillary Thyroid Microcarcinoma: 30-Year Experience
A concern with active surveillance is that some patients stop showing up for monitoring. A study using a Dutch cancer registry found that loss-to-follow-up rates were higher in the early period after diagnosis, though they stabilized after about two years. Reassuringly, no deaths were reported among patients who dropped out of surveillance.25PubMed Central. Active surveillance vs. surgery in low‐risk papillary thyroid microcarcinoma patients and the risk of loss to follow‐up Active surveillance is now endorsed by several major guidelines for appropriately selected low-risk microcarcinomas, though it requires a patient who is comfortable with the idea of living with a known cancer and committed to follow-up appointments. It is not appropriate for tumors near the trachea or recurrent laryngeal nerve, those with aggressive features on biopsy, or those with evidence of lymph node involvement.
For anyone facing a thyroid biopsy, the 5% malignancy figure is a reasonable starting expectation for the average patient. But the real answer depends on what your ultrasound looks like, which Bethesda category the pathologist assigns, your age, and increasingly, what molecular markers reveal. If the result comes back benign, the odds are strongly in your favor. If it comes back indeterminate, the path forward may involve more testing, but the majority of those nodules also turn out to be non-cancerous. And even among the cancers that are found, most are slow-growing papillary types where the prognosis, whether treated with surgery or monitored carefully, remains excellent.