Exemestane vs. Anastrozole: A Comparison of Efficacy

Exemestane and anastrozole perform similarly in the setting where they are most commonly prescribed: early-stage, hormone receptor-positive breast cancer in postmenopausal women. The largest head-to-head trial, MA.27, found identical four-year disease-free survival of about 91% for both drugs. But “similarly” is not “identically,” and the two drugs differ in meaningful ways beneath the surface, from how they disable the aromatase enzyme to how their byproducts interact with other hormone receptors. Those differences start to matter when the conversation moves beyond average outcomes and into bone health, cardiovascular risk, drug sequencing after resistance, and emerging pharmacogenomic data.

Two Ways to Shut Down the Same Enzyme

Both exemestane and anastrozole work by blocking aromatase, the enzyme responsible for converting androgens into estrogen in postmenopausal women. Without that conversion, estrogen levels plummet, starving hormone receptor-positive breast cancer cells of the fuel they need to grow. But the two drugs accomplish this through fundamentally different chemistry.

Anastrozole is a nonsteroidal aromatase inhibitor. It binds to the active site of the aromatase enzyme reversibly, meaning it sits in the way and blocks the enzyme’s function, but the enzyme itself remains intact. If the drug is withdrawn, aromatase activity can recover. Exemestane, by contrast, is a steroidal aromatase inactivator. Its structure mimics the enzyme’s natural substrate, so it slips into the active site and forms a permanent covalent bond, destroying that copy of the enzyme for good. New aromatase molecules must be produced from scratch before estrogen synthesis can resume.

This distinction has a practical consequence: exemestane generates a metabolite called 17-hydroexemestane that binds strongly to the androgen receptor. Lab studies show that 17-hydroexemestane latches onto the androgen receptor with an affinity close to that of the body’s own potent androgen, DHT.¹ This metabolite essentially acts as a mild androgen in the body.² Anastrozole produces no such androgenic byproduct. That androgenic activity is part of the rationale for why exemestane might protect bone slightly better and why it can sometimes work after a nonsteroidal inhibitor has failed, both of which come up later in this article.

The MA.27 Trial and Early Breast Cancer

The most important piece of evidence in this comparison is the NCIC CTG MA.27 trial, a large randomized phase III study that directly compared exemestane and anastrozole as upfront adjuvant therapy in postmenopausal women with hormone receptor-positive early breast cancer. It enrolled over 7,500 patients and followed them for a median of just over four years.

The result was unambiguous: the four-year event-free survival was 91% for exemestane and 91.2% for anastrozole, with a hazard ratio of 1.02 and no statistical difference between the groups. Distant recurrence, new breast cancers, and deaths all occurred at virtually the same rate in both arms. Overall survival and disease-specific survival were also similar.³ The MA.27 investigators had actually designed the trial hoping to show exemestane was superior. It was not. The survival curves sat on top of each other for the entire follow-up period.⁴

For the majority of patients making treatment decisions about early-stage disease, this trial remains the most clinically relevant piece of the puzzle. Both drugs are effective, and at the four-year mark, choosing one over the other based on cancer outcomes alone is not supported by the evidence.

What Longer Follow-Up Suggests

Four years of follow-up, while meaningful, is relatively short for a cancer that can recur a decade or more after initial treatment. A 2025 observational study published in JAMA Network Open looked at outcomes over a longer horizon, with a median follow-up of just over five years, and estimated eight-year survival rates. This study included patients treated with all three aromatase inhibitors: anastrozole, letrozole, and exemestane.

At eight years, disease-free survival was estimated at about 79% for exemestane, compared with roughly 81% for both anastrozole and letrozole. Overall survival followed a similar pattern, with exemestane at about 89% and anastrozole at roughly 90.5%.⁵ The gaps are small in absolute terms, but they are worth noting.

There are important caveats. This was an observational study, not a randomized trial, so differences in the types of patients prescribed each drug could account for some of the gap. Exemestane is sometimes selected for patients considered higher risk or for those who have already tried and stopped a nonsteroidal inhibitor, which would bias its outcomes downward. Still, the study raises a reasonable question about whether the equivalence seen at four years in MA.27 fully holds over a longer timeline. At this point, no randomized data with eight-year follow-up exists to settle the question definitively.

Advanced and Metastatic Breast Cancer

In the advanced or metastatic setting, the comparison between exemestane and anastrozole gets muddier, and the trial data is thinner. Two smaller phase II studies have looked at this directly, with somewhat divergent results.

A Spanish Breast Cancer Group trial randomized postmenopausal patients with hormone receptor-positive advanced breast cancer to receive either exemestane or anastrozole as first-line endocrine therapy. Anastrozole appeared to perform better, with an overall response rate of 46% versus about 36% for exemestane and a median time to progression of 12.1 months versus 6.1 months.⁶ Those numbers look meaningfully different, but this was a small phase II study, not a large definitive trial.

A Japanese randomized study, on the other hand, found similar time-to-progression between the two drugs. Exemestane showed a median time to progression of about 14 months compared with around 11 to 14 months for anastrozole depending on assessment method, with overlapping confidence intervals.⁷ The evidence here is genuinely mixed. The honest take is that no large, adequately powered trial has established whether one drug is clearly better than the other for advanced disease.

Effects on Bone

All aromatase inhibitors accelerate bone loss because estrogen is important for maintaining bone density. One of the most persistent questions in this comparison is whether exemestane’s androgenic metabolite gives it a protective edge over anastrozole when it comes to skeletal health.

The answer, based on actual trial data, is: the difference is smaller than the theory predicts. A companion bone mineral density substudy of the MA.27 trial measured changes in hip and spine density over two years. In patients with normal or mildly reduced bone density at baseline, exemestane users lost slightly less bone than anastrozole users at both the hip and spine, but the differences did not reach statistical significance.⁸ A Japanese substudy of the TEAM trial similarly reported comparable bone effects between the two drugs.⁹

In the MA.27B substudy, fracture counts were also tracked. Among patients starting with healthier bones, those on anastrozole had slightly more fractures than those on exemestane, but the numbers were very small and not powered for a definitive comparison. For patients who already had reduced bone density at baseline, both groups actually gained bone during the study, likely because those patients were receiving bisphosphonate therapy or other bone-protective interventions as part of their standard care.

The bottom line on bone: there is a theoretical reason exemestane might be gentler on the skeleton, backed by some laboratory and preclinical evidence related to its androgenic metabolite. In clinical practice, the difference has not clearly materialized in the available trial data. Both drugs warrant bone density monitoring and protective measures when needed.

Cardiovascular and Lipid Profiles

Cardiovascular risk is another area where the two drugs’ different mechanisms might theoretically lead to different side effect profiles. Estrogen has complex effects on blood vessels and lipids, and suppressing it differently might produce different downstream cardiovascular consequences.

A randomized study comparing the lipid effects of anastrozole, letrozole, and exemestane found that exemestane significantly lowered total cholesterol, which sounds beneficial, but it also significantly lowered HDL cholesterol, the “good” cholesterol. When the three drugs were compared head to head, no significant differences emerged between them for total cholesterol, LDL, HDL, or apolipoprotein B.¹⁰ So while exemestane’s effect on individual lipid components was detectable, the net cardiovascular lipid picture was not meaningfully different from anastrozole’s.

A network meta-analysis that pooled data from multiple trials estimated that the overall cardiovascular risk ranking, from highest to lowest, was letrozole, then exemestane, then anastrozole. But the confidence intervals for the comparisons between exemestane and anastrozole were wide and crossed 1.0, meaning the difference was not statistically significant.¹¹ In practice, for most patients, cardiovascular risk does not clearly favor one drug over the other. That said, for a patient with a pre-existing lipid disorder or established cardiovascular disease, an oncologist might factor these subtle trends into the decision.

Side Effects and Why People Stop Treatment

One of the biggest real-world challenges with aromatase inhibitors is not which one is marginally more effective but whether patients can tolerate taking them for the recommended five or more years. Joint pain, hot flashes, fatigue, mood changes, and musculoskeletal stiffness are common with all aromatase inhibitors and frequently lead patients to stop treatment early.

The MA.27 trial included a quality-of-life companion study that tracked patient-reported symptoms on both drugs. Treatment-related symptom scores were not significantly different between exemestane and anastrozole at any measured time point, including 3, 6, 12, and 24 months. Overall health-related quality of life also remained stable and similar in both arms.¹²

Discontinuation rates were high regardless of which drug patients received. About 20% of patients had stopped their aromatase inhibitor by two years, and roughly a third had stopped by four years. The strongest predictor of early discontinuation was not which drug a patient was assigned to but rather how much side-effect bother the patient reported even before starting treatment. Patients who were already troubled by symptoms at baseline were about 29% more likely to quit early.¹² This is a useful clinical insight: a patient’s baseline symptom burden may be a better guide to whether she will complete therapy than the specific aromatase inhibitor chosen.

Still, individual tolerance varies. Some patients who cannot stand one aromatase inhibitor find the other more manageable, even though the drugs produce similar average symptom profiles across large groups. Switching between a steroidal and nonsteroidal inhibitor is a common and reasonable clinical strategy when side effects become intolerable.

Using One After the Other Fails

The structural difference between exemestane and anastrozole becomes clinically significant in a scenario that unfortunately affects many patients: disease progression on one of the drugs. Because exemestane permanently inactivates aromatase while anastrozole only reversibly inhibits it, and because their molecular structures interact with the enzyme differently, partial cross-resistance between the two classes is not guaranteed. This opens the door to sequencing strategies.

A study of patients with advanced breast cancer who had progressed on a nonsteroidal aromatase inhibitor (anastrozole or letrozole) found that switching to exemestane yielded a clinical benefit rate of about 44%, with a median time to further progression of roughly five months.¹³ A separate phase II trial similarly documented a 44% clinical benefit rate with exemestane after progression on anastrozole, with evidence that cross-resistance between the two classes was not complete.¹⁴

A clinical benefit rate of 44% in patients whose cancer has already outsmarted one aromatase inhibitor is clinically meaningful. It does not represent cures, but it represents months of disease control that can delay the need for chemotherapy. This sequencing option is one of the practical advantages of having two structurally different aromatase inhibitors available, and it is one reason oncologists sometimes keep one in reserve rather than treating the drugs as fully interchangeable.

How Food and Formulation Affect Exemestane

A practical difference that often goes unmentioned in efficacy comparisons is that exemestane’s absorption is meaningfully affected by food. Pharmacokinetic modeling shows that exemestane absorbs differently depending on whether it is taken in a fasted or fed state and on the specific formulation used.¹⁵ The standard recommendation is to take exemestane after a meal, as this increases drug exposure. Anastrozole, by contrast, can be taken with or without food, making it somewhat simpler from a daily routine standpoint. This difference is unlikely to change cancer outcomes in a carefully monitored trial, but in the real world, where adherence is already a challenge, the extra instruction to always take exemestane with food adds one more thing to get right.

Pharmacogenomics and the Future of Choosing Between Them

If the two drugs produce similar outcomes on average, could genetic testing identify patients who would do better on one versus the other? This is an active area of research. An analysis of genetic data from over 4,400 patients in the MA.27 trial identified a group of genetic variants that were differently associated with outcomes depending on whether the patient was taking anastrozole or exemestane.¹⁶

The findings are preliminary and not yet ready for clinical use. No validated test exists today that tells a patient “you should take exemestane, not anastrozole, based on your genome.” But the concept is promising. Because the two drugs are metabolized through different pathways and have different active metabolites, it is biologically plausible that genetic variation in those metabolic enzymes could meaningfully shift which drug works better for a given individual. If this line of research matures, the current approach of treating these drugs as interchangeable could eventually be replaced by genotype-guided selection. For now, the choice between them remains a clinical judgment call based on side effect profiles, sequencing plans, patient preference, and individual risk factors rather than a precision-medicine algorithm.

Cost and Generic Availability

Both exemestane and anastrozole have been available as generics for well over a decade, and in most markets the out-of-pocket cost difference between them is minimal. This was not always the case. When the drugs were still on patent, cost-effectiveness analyses struggled to compare them because no direct head-to-head trial data existed at the time to anchor the models. Now that generics are widely available and the MA.27 trial has established clinical equivalence, cost rarely tips the scale decisively in either direction. Insurance formularies and regional availability are more likely to determine which drug a patient receives than any consistent price advantage for one over the other.