Letrozole, marketed as Femara, inhibits aromatase and reduces oestrogen production. Its role in clinical research changes with the trial. In Edinburgh breast studies, investigators used it to track how a tumour shrank over time before surgery. In ovarian research, they used it to ask whether the amount of oestrogen receptor in tumour tissue predicted disease control. More recently, two randomised ovarian studies used letrozole to answer different questions about targeted treatment and whether chemotherapy could be omitted. Those studies involved different disease stages, comparators and endpoints. 1, 3, 8, 10, 13
The Scottish work spans laboratory measurements and clinical trial coordination. The Edinburgh Breast Unit published serial response data from extended presurgical treatment. The University of Edinburgh team also ran small ovarian cancer studies that compared tumour tissue findings with blood markers and scans. In GOG 281/LOGS, the University of Glasgow repository records the Glasgow Clinical Trials Unit's coordination role in the UK. The trial included letrozole among the options available to investigators in its standard-care arm. 3, 8, 9, 11
Mechanism and pharmacology in endocrine trial design
In postmenopausal patients, aromatase converts androgens to oestrogens in peripheral tissues. Letrozole inhibits this enzyme, reducing a signal that can drive the growth of hormone-sensitive tumours. A change in circulating hormone levels establishes an effect on the drug's target. A change in growth rate on a biopsy or scan asks whether the tumour responds to that effect. Eventually, a study may measure whether treatment delays recurrence or progression. Those measurements occur on different timescales and require different samples and follow-up. 1, 2, 3, 6
The Edinburgh crossover study followed 54 patients through periods on both anastrozole and letrozole. Using the same patients for both treatments reduced variability from differences in baseline hormone levels. After letrozole, oestradiol remained at or above the assay's 3 pmol/L threshold in one of 54 patients. After anastrozole, it did so in 20 of 54. The investigators also measured oestrone sulphate and found lower residual levels with letrozole, with levels at 2.0% of baseline compared with 4.6% on anastrozole. These are biochemical comparisons. The study design and sample size did not address comparative recurrence or survival. 2
At low concentrations, the assay's sensitivity can affect how often it reports a measurable result. A protocol measuring suppression should state the baseline, the lowest reliably measured concentration and how it handles results below that range. A trial comparing the two agents for clinical benefit needs an endpoint such as recurrence, where more complete hormone suppression is a possible explanation for a finding. Evidence for that endpoint comes from follow-up, rather than from the enzyme measurement alone. 2, 6
Breast oncology - how long tumours keep responding before surgery
The Edinburgh Breast Unit reported on 182 consecutive patients who received presurgical letrozole for at least three months. At three months, 69.8% had a partial or complete clinical response. Sixty-three patients continued beyond that point, and the authors reported that the overall response rate reached 83.5% with prolonged treatment. The team also tracked the operation that became possible. Among 134 patients initially assessed as requiring mastectomy, 81 were suitable for breast conservation at three months and 96 became suitable with extended treatment. Those figures describe a change in surgical eligibility in the cohort. 3
Clinicians chose who would continue, taking account of the response and individual circumstances. The comparison between three months and longer treatment therefore reflects clinical selection as well as time on treatment. The principal contribution is the observation that some tumours continued shrinking after the first assessment. A study that stops measuring at three months cannot describe that later course. To estimate the effect of assigning a later operation, investigators would need a design that accounted for early responders, patients who progressed and those who went directly to surgery. 3
A prospective multicentre study later followed patients who were initially unsuitable for breast-conserving surgery. Of 146 patients enrolled, 139 had a valid post-baseline tumour assessment and 96 of those 139 became eligible for breast conservation. The estimated median time to a sufficient response was 7.5 months (95% confidence interval [CI] 6.3-8.5 months). The endpoint was eligibility rather than the operation actually performed. The trial supplied a time-to-response estimate for a selected population, with treatment ending if the cancer progressed, a surgical decision was made or the participant reached another stopping criterion. 4
NeoPAL illustrates why an investigator may pair clinical response with a tissue endpoint. In 106 patients with high-risk luminal breast cancer, the phase II study assigned letrozole plus palbociclib or chemotherapy before surgery. At operation, 7.7% in the endocrine combination group and 15.7% in the chemotherapy group had little or no residual cancer by the trial's residual cancer burden categories. The trial reported broadly similar clinical response and breast conservation figures. Tumour size on examination versus residual cancer in surgical tissue gives two different accounts of response. The small study did not establish that the two strategies have equivalent long-term outcomes. 5
Presurgical studies - response, timing and residual disease
| Study | Population and design | Endpoint | Published result |
|---|---|---|---|
| Edinburgh crossover study | 54 postmenopausal patients received both aromatase inhibitors | Oestradiol at or above 3 pmol/L | 1/54 after letrozole versus 20/54 after anastrozole. |
| Edinburgh presurgical cohort | 182 consecutive patients, 63 continued beyond three months | Clinical response | 69.8% at three months and 83.5% after extended treatment in the published cohort analysis. |
| Edinburgh surgical eligibility | 134 patients initially assessed as requiring mastectomy | Suitability for breast conservation | 81/134 at three months, rising to 96/134 with further treatment. |
| Prospective duration study | 139 assessable participants initially unsuitable for breast conservation | Time to sufficient response to allow breast conservation | 96/139 became eligible. Estimated median time 7.5 months. |
| NeoPAL | 106 patients with high-risk luminal breast cancer in a phase II comparison | Residual cancer burden category 0 or I | 7.7% with letrozole plus palbociclib, 15.7% with chemotherapy. |
Adjuvant treatment - prior exposure changes the question
The Italian GIM4 trial asked what happens when patients who have already taken tamoxifen for two to three years receive a longer course of letrozole. It randomised 2,056 patients without recurrence to two to three years or five years of letrozole. After a median 11.7 years of follow-up, investigators reported estimated 12-year disease-free survival of 62% with the shorter letrozole course and 67% with the longer course. The hazard ratio was 0.78 (95% CI 0.65-0.93). Estimated 12-year overall survival was 84% and 88% respectively. The authors also reported more osteoporosis and joint symptoms with the longer course. 6
MA.17R started after roughly five years of aromatase inhibition, so participants entered with a different treatment history. It compared another five years of letrozole with placebo among 1,918 women who had already completed a substantial course. Five-year disease-free survival was 95% versus 91%, while overall survival showed no clear difference. Fractures and new osteoporosis occurred more often with further letrozole. GIM4 and MA.17R tested different stages of extended treatment and enrolled patients who had tolerated different earlier regimens. Prior exposure and trial eligibility help explain why a single summary of treatment duration would conceal clinically relevant differences. 6, 7
Gynaecological oncology - receptor expression and the Edinburgh studies
Bowman and colleagues studied letrozole in 60 patients with relapsed ovarian cancer and evaluated tumour markers in available tissue. Of 50 patients evaluable on imaging, none had a complete or partial response and ten had stable disease for at least 12 weeks. The blood-marker analysis had a separate denominator. Of 54 patients evaluable for CA125 response, five had a fall that met the response definition. Tumours with higher oestrogen receptor expression were associated with marker response and disease stabilisation. The results suggested a subgroup to study more closely and showed why scans and CA125 require separate reporting. 8
Smyth and colleagues then enrolled a population selected for oestrogen receptor expression. Seven of 42 patients evaluable for CA125 response had a decrease of more than 50%. On imaging, three of 33 had a partial response and 14 had stable disease at 12 weeks. In an exploratory analysis, CA125 response ranged from 0% in the immunoscore band of 150-199 to 33% in the band of 250-300. That gradient is an example of how tissue results can help design a later trial. A small single-arm study cannot establish a threshold that reliably predicts benefit in other populations. 9
The sample and assay matter. A block collected at the original operation describes a tumour before subsequent courses of treatment, while tissue sampled at recurrence may reflect a later state. A receptor score also depends on the laboratory method and its scoring rules. To turn the Edinburgh observation into a testable selection strategy, a later protocol would need to define how it obtains and reads tissue, which tumour types it includes, and whether it measures change by imaging, CA125, or both. Those decisions affect who can enter the study and what a reported response means. 8, 9
Glasgow coordination and translational work in GOG 281/LOGS
GOG 281/LOGS enrolled 260 patients with recurrent low-grade serous carcinoma. Before randomisation, investigators selected the standard-care option they would use if the patient entered the control group. The five options included letrozole, tamoxifen and three chemotherapy agents. The randomised arms had median progression-free survival of 13.0 months with trametinib and 7.2 months with standard care. In the control arm, six of 44 patients treated with letrozole had an objective response. That figure describes a group selected to receive letrozole within a broader comparison. The trial did not randomise letrozole against the other standard-care options. 10, 11
A 2026 exploratory publication examined tumour material from 170 trial participants. It investigated phosphorylated ERK, a marker of activity in the MAPK signalling pathway, and alterations in genes including KRAS, BRAF and NRAS. In the subgroup with high tumour phosphorylated ERK, median progression-free survival was 20.1 months with trametinib and 5.6 months with pooled standard care. The authors described this as a candidate marker of benefit from trametinib. Its role in the exploratory analysis differs from the oestrogen receptor measurements in the Edinburgh studies, which examined whether tumours responded to letrozole. 9, 12
NRG-GY019 - the first-line question after surgery
NRG-GY019 studied newly diagnosed stage II-IV low-grade serous carcinoma of the ovary or peritoneum after surgery. Both arms used letrozole. In one arm, patients received it after paclitaxel and carboplatin. In the other, they started letrozole without chemotherapy. This phase III trial asked whether omitting chemotherapy could preserve progression-free survival within an agreed margin. It tested a first-line strategy, while LOGS addressed treatment after recurrence. 10, 13
At the planned second interim analysis, presented in April 2026, median follow-up was 27.3 months. The hazard ratio for progression or death with letrozole alone versus chemotherapy followed by letrozole was 1.30 (95% CI 0.90-1.89). The study crossed its prespecified futility boundary and did not establish non-inferiority. Although the interval included 1, it also included values above the margin. The trial group reported substantially more severe adverse events with the chemotherapy strategy, a separate finding about treatment burden. These figures come from an interim analysis presented at a meeting, with the source identified accordingly. 13
MATAO - a separate question about maintenance
The MATAO protocol randomises maintenance letrozole against placebo after surgery and platinum-based chemotherapy. It enrols patients with newly diagnosed stage II-IV oestrogen-receptor-positive epithelial ovarian cancer, including low-grade or high-grade serous and endometrioid disease. Its receptor criterion requires at least 1% expression on central assessment. That threshold defines which patients can enter this trial. It does not carry the same meaning as the receptor-intensity bands in the earlier Edinburgh analysis, which looked for differences among selected patients with relapsed cancer. 9, 14
The placebo-controlled design isolates the role of maintenance after chemotherapy. In NRG-GY019 both groups received letrozole, which isolates the clinical strategy of adding chemotherapy before endocrine therapy. Those designs could produce results that matter to different clinical decisions at the same time. The MATAO source cited here is the published protocol. It establishes the entry criteria and planned progression-free survival comparison without supplying an efficacy estimate. 13, 14
Ovarian cancer studies - populations and endpoints
| Study | Clinical setting | Endpoint or comparison | Result and evidence stage |
|---|---|---|---|
| Bowman phase II | Relapsed ovarian cancer | CA125 response and imaging response | Five CA125 responses among 54 evaluable patients. No objective imaging responses among 50. |
| Smyth phase II | Oestrogen-receptor-positive relapsed ovarian cancer | CA125 response and radiological response | 7/42 CA125 responses and 3/33 partial responses on imaging. |
| GOG 281/LOGS | Recurrent low-grade serous carcinoma | Trametinib versus physician's choice | 13.0 versus 7.2 months median progression-free survival. Six responses among 44 letrozole recipients within the control group. |
| NRG-GY019 | Newly diagnosed stage II-IV low-grade serous carcinoma after surgery | Letrozole alone versus chemotherapy followed by letrozole | Second interim analysis hazard ratio 1.30 (95% CI 0.90-1.89). Non-inferiority was not established. |
| MATAO | After platinum-based treatment for ER-positive epithelial ovarian cancer | Maintenance letrozole versus placebo | Published phase III protocol. Progression-free survival is the planned primary endpoint. |
Safety, eligibility and reporting in extended treatment trials
A long-duration study needs to track patients who stop or switch treatment as well as those who finish as planned. Joint symptoms can lead to interruption, while changes in bone density or a fracture may prompt additional treatment. Reporting time on therapy and the reason for discontinuation alongside the recurrence endpoint lets readers see what patients actually experienced. GIM4 reported osteoporosis in 8% with the longer course and 5% with the shorter course. MA.17R also documented additional bone toxicity but studied women who had already completed earlier aromatase inhibition. 6, 7
Study populations can also differ in whether they enter before surgery, after adjuvant treatment or after a radiological relapse. That timing determines which baseline information needs to be collected. In a presurgical study, an investigator can record whether a patient was considered eligible for breast conservation at entry. In a long-term trial, the protocol may instead need a detailed endocrine treatment history and defined bone-health assessments. These data make it possible to interpret benefit and adverse effects in the group actually enrolled. 3, 4, 6, 7
Extended adjuvant studies - outcomes and trade-offs
| Study | Previous endocrine treatment | Disease outcome | Bone outcome |
|---|---|---|---|
| GIM4 | Two to three years of tamoxifen before randomisation | 12-year disease-free survival of 67% with the longer course and 62% with the shorter course. | Osteoporosis in 8% versus 5%, respectively. |
| MA.17R | Approximately five years of an aromatase inhibitor | Five-year disease-free survival of 95% with further letrozole and 91% with placebo. | More fractures, bone pain and new osteoporosis with further letrozole. |
References
- Femara UK summary of product characteristics.
- Dixon and colleagues, Journal of Clinical Oncology, 2008. Edinburgh crossover study of plasma oestrogen suppression.
- Dixon and colleagues, Breast Cancer Research and Treatment, 2009. Extended presurgical treatment in the Edinburgh cohort.
- Breast Cancer Research and Treatment, 2014. Prospective breast conservation eligibility study.
- Cottu and colleagues, Annals of Oncology, 2018. NeoPAL phase II trial.
- Del Mastro and colleagues, Lancet Oncology, 2021. GIM4 long-term results.
- Goss and colleagues, New England Journal of Medicine, 2016. MA.17R extension study.
- Bowman and colleagues, Clinical Cancer Research, 2002. Edinburgh ovarian phase II study with receptor analysis.
- Smyth and colleagues, Clinical Cancer Research, 2007. Study in oestrogen-receptor-positive relapsed ovarian cancer.
- Gershenson and colleagues, Lancet, 2022. Full GOG 281/LOGS report.
- University of Glasgow. GOG 281/LOGS research record and coordination information.
- Hollis and colleagues, Clinical Cancer Research, 2026. Tumour biomarker analysis from GOG 281/LOGS.
- NRG Oncology, April 2026. NRG-GY019 second interim analysis presented at the SGO annual meeting.
- McLaughlin and colleagues, BMC Cancer, 2022. MATAO phase III maintenance trial protocol.
