Afirma is a genomic test designed to help determine whether a thyroid nodule that looks uncertain under the microscope is likely benign or suspicious for cancer. When a fine-needle aspiration biopsy of a thyroid nodule comes back as “indeterminate,” meaning the cells do not look clearly benign or clearly malignant, Afirma analyzes gene expression patterns in the sample to guide the next step. The goal is straightforward: spare people with benign nodules from unnecessary thyroid surgery while still flagging those who need an operation. A large meta-analysis found that the current version of the test catches roughly 97% of cancers while correctly identifying the majority of benign nodules, though the real-world numbers depend on who is being tested and how the results are used.
Why Indeterminate Nodules Are a Problem
Thyroid nodules are common. Most are benign, but the only way to be sure is usually a biopsy. Pathologists examine the cells under a microscope and assign a category on the Bethesda scale, which runs from I (not enough cells to read) through VI (clearly malignant). The trouble zone is Bethesda III (atypia of undetermined significance) and Bethesda IV (follicular neoplasm or suspicious for one). In these categories, the cells look a bit “off” but not definitively cancerous. The cancer risk in these groups generally falls somewhere between about 15% and 30%, depending on the institution. Before molecular testing existed, many patients with these results went straight to surgery, and the majority turned out not to have cancer at all.
That is the gap Afirma was built to fill. If a genomic test can reliably say “this nodule is almost certainly benign,” many of those surgeries become avoidable. American Thyroid Association guidelines recommend considering molecular testing for Bethesda III and IV nodules when the result could change the management plan.1Endocrine Reviews. Diagnostic Utility of Molecular and Imaging Biomarkers in Cytological Indeterminate Thyroid Nodules
How the Test Works
Afirma analyzes RNA extracted from the same cells collected during a fine-needle aspiration biopsy, so no additional needle stick is needed as long as the sample was collected in the proper preservation solution. The test looks at gene expression patterns across thousands of genes using machine-learning algorithms trained on known benign and malignant nodule samples. The result comes back as either “benign” (suggesting active surveillance rather than surgery) or “suspicious” (suggesting that surgery should be considered). Analytical studies have shown the test is robust, tolerating variation in RNA input and remaining accurate even when the sample is diluted with up to 75% blood or contains up to 30% genomic DNA.2PubMed Central. Performance of Afirma Genomic Sequencing Classifier in the Diagnosis of Cytologically Indeterminate Thyroid Nodules
The test is fundamentally a “rule-out” tool. Its primary strength is its high negative predictive value: when Afirma says “benign,” you can be quite confident the nodule is not cancer. That confidence is what allows physicians and patients to choose monitoring over an operating room.
From GEC to GSC
Afirma has gone through a significant upgrade. The original version, called the Gene Expression Classifier (GEC), launched around 2012 and was effective at ruling out cancer but had a well-known weakness: too many benign nodules came back as “suspicious.” In technical terms, its specificity was low, meaning it flagged a lot of nodules that turned out to be harmless. The second-generation Genomic Sequencing Classifier (GSC) was developed specifically to improve that specificity.3PubMed. Real-world Comparison of Afirma GEC and GSC for the Assessment of Cytologically Indeterminate Thyroid Nodules
The improvement was substantial. A meta-analysis comparing the two generations found that the GSC’s specificity and positive predictive value were both significantly better than the GEC’s, while still maintaining high sensitivity. The GSC’s sensitivity was around 94%, its specificity rose to about 43% from the GEC’s 25%, and its positive predictive value climbed from roughly 42% to 63%.4Cancer Cytopathology. Diagnostic performances of the Afirma Gene Sequencing Classifier in comparison with the Gene Expression Classifier: A meta‐analysis In real-world practice, the GSC has performed even better than in its original validation study: a meta-analysis of real-world data found specificity around 80-88% and a negative predictive value above 99%.5PubMed Central. Real-World Performance of the Afirma Genomic Sequencing Classifier (GSC)—A Meta-analysis
The Hürthle Cell Problem
One of the most significant improvements in the GSC upgrade involved Hürthle cell nodules, which have been a persistent headache in thyroid diagnostics. These cells have unusual chromosomal features that confused the original GEC, leading it to call the vast majority of Hürthle cell nodules “suspicious” even when they were benign. With the GEC, specificity for Hürthle cell neoplasms was just 12%. The GSC, which uses a dedicated set of algorithms for these tumors, brought that up to about 59% while maintaining the same sensitivity of 89%.6PubMed Central. Identification of hürthle cell cancers: solving a clinical challenge with genomic sequencing and a trio of machine learning algorithms
An independent comparison confirmed this in clinical practice. For Bethesda IV nodules suspicious for Hürthle cell neoplasm, the GSC’s benign call rate was 68%, compared to just 16% with the old GEC.7Thyroid®. Independent Comparison of the Afirma Genomic Sequencing Classifier and Gene Expression Classifier for Cytologically Indeterminate Thyroid Nodules That is a dramatic shift for patients who previously would have been sent to surgery almost by default because the older test could not adequately evaluate their nodule type.
Does It Actually Reduce Unnecessary Surgery?
In theory, a benign Afirma result should keep people out of the operating room. The evidence mostly supports this, though the picture is not perfectly clean. One cost-effectiveness analysis estimated that Afirma reduced the probability of unnecessary surgery from about 76% to roughly 37%, at a cost of around $8,400 per unnecessary surgery avoided.8PubMed Central. Cost-Effectiveness of Lobectomy Versus Genetic Testing (Afirma) for Indeterminate Thyroid Nodules: Considering the Costs of Surveillance A long-term follow-up study found that a benign GEC result was associated with a durable reduction in the rate of surgery, with most of the few operations happening in the first year and very few beyond two years.9Endocrine Practice. Long-Term Nonoperative Rate of Thyroid Nodules with Benign Results on the Afirma Gene Expression Classifier
But one study sounded a cautionary note. It found that the rate of surgery did not change significantly after Afirma was introduced at one institution, and suggested that the test might have an unintended effect of increasing the number of nodules called “indeterminate” in the first place, without ultimately lowering the thyroidectomy rate.10Cancer Cytopathology. Impact of Afirma gene expression classifier on cytopathology diagnosis and rate of thyroidectomy This is a reminder that how the test is integrated into a practice’s workflow matters as much as the test itself. When physicians and pathologists trust the result and adjust their behavior accordingly, the benefits show up. When the result is overridden by clinical anxiety or institutional culture, the test may add cost without changing outcomes.
A more recent retrospective study with both Bethesda III and IV nodules found that patients who underwent molecular testing received optimal surgical management (meaning the right operation for the right diagnosis) about 68% of the time, compared to 25% for those who did not get molecular testing.11PubMed Central. Molecular Testing and Surgical Outcomes in Bethesda III and IV Thyroid Nodules: A Retrospective Cohort Study The benefit was consistent across both Bethesda categories.
How Afirma Compares to ThyroSeq
Afirma is not the only molecular test for thyroid nodules. ThyroSeq, a DNA and RNA-based next-generation sequencing panel, is its main competitor. The two tests take different technical approaches. Afirma classifies the overall gene expression pattern of a nodule, while ThyroSeq looks for specific mutations and gene fusions known to be associated with thyroid cancer. Patients often want to know which one is “better,” and the honest answer from the literature is that they perform similarly for most patients.
A systematic review and meta-analysis comparing the second-generation versions of both tests found no statistically significant difference in sensitivity, specificity, positive predictive value, or negative predictive value between Afirma GSC and ThyroSeq v3.12PubMed. Diagnostic performance of the second-generation molecular tests in the assessment of indeterminate thyroid nodules: A systematic review and meta-analysis A head-to-head study at a single institution found that both tests had high sensitivity (100% for Afirma GSC, 97% for ThyroSeq v3) and comparable specificity (77% vs. 83%), with the differences not reaching statistical significance.13The Journal of Clinical Endocrinology & Metabolism. Bethesda III and IV Thyroid Nodules Managed Nonoperatively After Molecular Testing With Afirma GSC or Thyroseq v3
A separate meta-analysis that used area-under-the-curve analysis did find ThyroSeq v3 had the best overall discriminatory performance (AUC 0.95) compared to Afirma GSC (AUC 0.90), and suggested ThyroSeq might have somewhat better “rule-in” properties, meaning its suspicious results carry a higher probability of actual malignancy.14PubMed Central. Thyroseq v3, Afirma GSC, and microRNA Panels Versus Previous Molecular Tests in the Preoperative Diagnosis of Indeterminate Thyroid Nodules: A Systematic Review and Meta-Analysis In practice, which test your endocrinologist or surgeon orders often depends on institutional preference, laboratory contracts, and which test’s reporting format they find most useful, rather than a clear superiority of one over the other.
What a Benign Result Means for Follow-Up
A benign Afirma result does not mean you can forget about the nodule entirely. Current practice calls for continued ultrasound surveillance, typically at regular intervals, to watch for any growth or change in appearance that might warrant repeat biopsy or eventual surgery. In a study that followed patients with benign or negative molecular test results for a median of nearly three years, about 92% were managed without surgery. Among the 15 nodules that were eventually resected during surveillance, only one turned out to be malignant, and that one had tested negative on ThyroSeq, not Afirma. The overall false-negative rate across both tests was less than 1%.15The Journal of Clinical Endocrinology & Metabolism. Bethesda III and IV Thyroid Nodules Managed Nonoperatively After Molecular Testing With Afirma GSC or Thyroseq v3
That said, an earlier study using the older-generation tests (Afirma GEC and ThyroSeq v2) found a higher false-negative rate of about 6% across both platforms.16The Journal of Clinical Endocrinology & Metabolism. Outcomes of Indeterminate Thyroid Nodules Managed Nonoperatively after Molecular Testing The takeaway is that the current-generation tests appear substantially safer than their predecessors, but continued monitoring is still the standard of care after a benign molecular result. No test is perfect, and a small number of cancers will be missed. Most of those missed cancers, when they are found during surveillance, are low-risk and treatable.
When Afirma Might Be Less Reliable
Afirma’s performance is not uniform across every clinical scenario. One known limitation is that the test’s negative predictive value depends on the underlying cancer prevalence in the population being tested. At institutions where indeterminate nodules turn out to be cancerous at higher-than-average rates, a benign Afirma result is inherently less reassuring. A study from a practice where 33% of indeterminate biopsies proved malignant found the negative predictive value of the GEC was not as robust as the published literature suggested.17PubMed. Surgical utility of Afirma: effects of high cancer prevalence and oncocytic cell types in patients with indeterminate thyroid cytology This is a basic statistical reality that applies to any diagnostic test: the more common the disease, the harder it is for a negative result to fully rule it out. Patients at referral centers with high cancer concentrations may need to weigh this factor.
Performance in Younger Patients
Thyroid cancer in children and young adults raises particular anxiety. These patients often have higher rates of more aggressive cancer subtypes, and the research base for molecular testing has been overwhelmingly adult. A recent study specifically evaluated Afirma GSC in young patients with indeterminate thyroid nodules and found that the test caught all 14 malignancies (sensitivity of 100%) and correctly called 30 of 35 benign cases as benign (specificity of 86%). The negative predictive value was 100%, meaning no cancers were missed among nodules called benign.18The Journal of Clinical Endocrinology & Metabolism. Afirma genomic sequencing classifier performance in young patients with cytologically indeterminate thyroid nodules While the sample size was small, these are encouraging numbers and suggest the test works in a younger population, though more data is needed before anyone can be fully confident.
The Xpression Atlas and Surgical Planning
When Afirma classifies a nodule as suspicious, the story does not end with a binary yes-or-no. The Afirma Xpression Atlas (XA) is a companion report that identifies specific genomic alterations in the suspicious sample. Among over 16,000 suspicious indeterminate nodules, about 44% had at least one identifiable genomic alteration on the expanded XA panel.19PubMed Central. Afirma Genomic Sequencing Classifier and Xpression Atlas Molecular Findings in Consecutive Bethesda III-VI Thyroid Nodules
This matters because the specific alteration can influence what kind of surgery is appropriate, whether additional preoperative testing is needed, and even whether certain hereditary syndromes should be investigated. For example, finding a BRAF V600E mutation might push toward a more aggressive surgical approach, while a RAS mutation might support a lobectomy. The XA report can also flag potential targets for systemic therapy if the cancer turns out to be advanced.20Cancer Cytopathology. The Afirma Xpression Atlas for thyroid nodules and thyroid cancer metastases: Insights to inform clinical decision‐making from a fine‐needle aspiration sample The shift from a simple benign-or-suspicious binary toward a more detailed molecular portrait is where thyroid diagnostics are clearly headed.
The BRAF-Like and RAS-Like Distinction
Understanding why genomic profiling matters requires a brief look at how thyroid cancers differ at the molecular level. Research from large-scale projects has shown that papillary thyroid cancers generally fall into two broad molecular subtypes based on their gene expression patterns: “BRAF-like” and “RAS-like.” These subtypes predict biological behavior more accurately than looking at the cancer under a microscope alone. BRAF-like tumors tend to be more aggressive, less responsive to radioactive iodine treatment, and enriched in signaling pathways that drive cell growth. RAS-like tumors tend to behave more indolently.21PubMed Central. Leveraging the transcriptome-phenotype relationship to guide clinical management of papillary thyroid cancer
The picture can be more complex than a simple split, though. Some follicular-variant papillary cancers with a BRAF mutation still cluster with RAS-like tumors on gene expression analysis, while others with the same mutation look classically BRAF-like and behave more aggressively, particularly when they grow in an infiltrative rather than an encapsulated pattern.22PubMed. Comprehensive Transcriptomic and Genomic Profiling of Subtypes of Follicular Variant of Papillary Thyroid Carcinoma This complexity is precisely why gene expression-based classification adds value on top of just looking for a single mutation. Newer tools are being developed that use these transcriptomic subtypes to guide decisions about radioactive iodine therapy, the extent of surgery, and long-term surveillance intensity.
Cost and Insurance Considerations
Molecular testing for thyroid nodules is not cheap. The Afirma test typically runs several thousand dollars, and coverage varies by insurer. For patients weighing the value, the relevant question is whether the cost of testing is less than the cost (financial and physical) of an unnecessary thyroid surgery, including the lifelong thyroid hormone replacement that comes with a total thyroidectomy. A cost-effectiveness analysis found that using Afirma for indeterminate solitary thyroid nodules was a cost-effective strategy for avoiding unnecessary surgery. At a willingness-to-pay threshold of $5,000, molecular testing had a 63% chance of being the more cost-effective approach compared to proceeding directly to surgery.23Journal of Otolaryngology – Head & Neck Surgery. Cost-effectiveness analysis of molecular testing for cytologically indeterminate thyroid nodules
The economics shift when you factor in the alternative of long-term surveillance alone without molecular testing. If a patient and their physician are comfortable monitoring an indeterminate nodule indefinitely with periodic ultrasounds and possible repeat biopsies, the upfront cost of the molecular test might not be justified. But for many people, the psychological burden of living with an uncertain diagnosis is itself a real cost. Molecular testing often resolves that uncertainty one way or the other, which has value that does not show up in standard cost-effectiveness models.
Where the Technology Is Heading
Afirma and its competitors have been improving rapidly. The jump from the GEC to the GSC was not just an incremental tweak; it was a fundamental rearchitecting of the classifier that incorporated RNA sequencing and new machine-learning approaches. The addition of the Xpression Atlas turned a binary benign-or-suspicious test into something that could also characterize the specific molecular drivers of a suspicious nodule. Newer transcriptomic tools are pushing the concept further, attempting to use gene expression subtypes to predict not just whether a nodule is cancer, but how that cancer will behave over time, whether it will respond to radioactive iodine, and whether a given patient needs aggressive treatment or can be managed conservatively.
The broader direction is a move away from one-size-fits-all surgical management and toward biologically informed decisions. A decade ago, an indeterminate thyroid nodule often meant a diagnostic lobectomy just to find out what was going on. Today, molecular testing spares many patients from that surgery entirely. For those who do need surgery, the genomic information available preoperatively is increasingly useful in planning how much thyroid to remove and what additional workup to pursue. The era of “operate first, diagnose second” for ambiguous thyroid nodules is not over, but it is shrinking.