Part XI · Hormone receptor-positive, HER2-negative disease · Chapter 63
Endocrine resistance, and the presentations the sequence does not fit
How to order the branches of this disease, and why almost every trial here randomised something other than the order.
1 · Defining endocrine sensitivity and resistance in clinical language
Endocrine resistance is a clinical label applied to a heterogeneous set of failures. The consensus definitions divide it by timing. Primary resistance means relapse within the first two years of adjuvant endocrine therapy, or progression within six months of starting endocrine therapy for advanced disease. Secondary resistance means failure after a longer period of benefit.
These are useful conventions and they are conventions. Nobody has shown that six months marks a point where biology changes. The cut-points exist so that trial populations can be defined reproducibly, which is the same argument made about assay thresholds in HER2 heterogeneity. Reading a trial eligibility boundary as a biological category is the recurring error in this literature.
What the timing does encode is real, and it is worth naming. A tumour that never responded probably never depended on oestrogen receptor signalling in a way the drug could exploit. A tumour that responded for four years and then failed has produced something. The first situation argues for changing class. The second argues for characterising what changed before choosing. The mechanisms behind each are separated in Endocrine resistance, ESR1 alterations, ER loss, and bypass signaling and the pathway detail is in Estrogen receptor signaling and endocrine resistance.
One practical consequence follows. The single most informative number at progression is how long the last line worked. It is free and it is already in the chart. It tells you which of the two situations above you are in, which no feature of the scan does.
This chapter is about how to approach the patient. Its subjects are disease burden, visceral crisis, endocrine sensitivity, the goal of each line, and when to stop. It is not the trial record for this disease, and does not try to be.
The tables below are generated from the trial registry when the book is built. Each is filtered on the decision its own section argues rather than on a drug class. A figure stays in the prose only where the sentence is about that figure. Everything else has been left to the table, so a new readout enters the registry once and appears wherever its filter matches.
2 · Rapid progression and visceral crisis
Visceral crisis is a clinical diagnosis and it is overused. It means organ dysfunction from tumour burden that requires a rapid response, judged by symptoms, examination and laboratory values. The international consensus guidelines put it the same way: not the mere presence of visceral metastases, but important organ compromise that makes the most rapidly efficacious therapy clinically necessary1,2. Extensive liver metastases with normal liver function are not a visceral crisis. They are a large burden.
The distinction matters because the reflex answer, combination chemotherapy, is not supported in the population where it is most often used.
Advanced disease · HR+/HER2- · 1L
Choosing the first-line regimen and the moment to leave it
Can endocrine therapy with a CDK4/6 inhibitor replace chemotherapy when the disease is moving fast?
Rationale
Aggressive visceral disease is where clinicians reach for chemotherapy first, on the belief that only cytotoxic treatment produces a response quickly enough to matter. That belief was formed against the slow and partial responses of endocrine monotherapy. The combination is not endocrine monotherapy, which is reason enough to test the belief rather than carry it forward.
Experiment
A randomised comparison of ribociclib with endocrine therapy against investigator's choice combination chemotherapy, in first-line disease the investigator judged aggressive enough to warrant chemotherapy. Alongside it, two trials set an inhibitor with endocrine therapy against capecitabine, one in premenopausal patients with ovarian suppression and one in aromatase inhibitor-resistant disease. A further trial randomises immediate abemaciclib with endocrine therapy against a short course of paclitaxel given before the same combination.
Finding
In aggressive first-line disease the endocrine combination held the disease longer than combination chemotherapy did. Against capecitabine the answer depends on the population. In premenopausal patients with ovarian suppression the combination lengthened time to progression against capecitabine. In aromatase inhibitor-resistant disease it did not beat capecitabine.
Limitation
The aggressive-disease trial was a phase II. Its comparator was a combination chemotherapy regimen rather than the sequential single agents many clinics would use, which is a comparator chosen for toxicity as much as for efficacy. Its endpoint was progression-free survival and not the speed of response on which the practice it challenges actually rests. Entry turned on the investigator's judgement that chemotherapy was indicated rather than on a stated visceral crisis threshold. That makes the population a product of a decision rather than of a measurement, which no protocol can reproduce. Of the two capecitabine trials, one enrolled a population only half of whom were untreated for metastatic disease. The other was not first line at all. The induction trial has no verified peer-reviewed result on the record. Nothing about giving chemotherapy before the combination can be read from it yet. An early response to induction would in any case not be an answer for every urgent presentation. Its eligibility and its exact sequence have to travel with it whenever it is cited.
Next question
If the endocrine combination can carry disease this aggressive, is an aromatase inhibitor still the right endocrine half of it?
RIGHT ChoiceYoung-PEARL / KCSG-BR15-10PEARLABIGAIL
RIGHT Choice was the first randomised test of that reflex, and it tests it narrowly. It is a phase II trial. It enrolled pre- and perimenopausal women only. Entry turned on the investigator's judgement that chemotherapy was indicated, which is a decision rather than a measurement, and no protocol reproduces it. Its comparator was combination chemotherapy, chosen for toxicity contrast as much as for efficacy, rather than the sequential single agents most clinics would give. Its endpoint was progression-free survival3. The practice it challenges does not rest on progression-free survival. It rests on how quickly the disease turns around.
A second randomised trial answers two of those objections, the menopausal restriction and the combination comparator. PADMA randomised 120 patients with an indication for chemotherapy, 88.3% of them postmenopausal, to palbociclib with endocrine therapy or to single-agent chemotherapy, mostly capecitabine or paclitaxel4. Median time to treatment failure, its primary endpoint, was 17.2 months against 6.1 months, with a hazard ratio of 0.46. Median progression-free survival was 18.7 months against 7.8 months, and overall survival did not differ significantly. It was a phase IV trial of 120 patients, and entry still turned on an indication for chemotherapy. Its endpoints measured how long treatment lasted rather than how quickly the disease turned around.
The two capecitabine trials in the story are not about urgency at all. PEARL and Young-PEARL / KCSG-BR15-10 enrolled on endocrine resistance and on menopausal status, and they are argued in Single-agent chemotherapy and when it remains the right choice. They matter here only in showing an endocrine combination holding disease as long as a cytotoxic agent in one population and not in another. That is a reason to ask which population is in front of you. It is not a licence to generalise from either.
ABIGAIL is the only design that addresses the manoeuvre clinicians actually use, which is giving chemotherapy first and handing over to the combination. It randomises immediate abemaciclib with endocrine therapy against a short course of paclitaxel before the same combination. It has no verified peer-reviewed result on the record, so nothing about induction can be read from it yet.
Caution
RIGHT Choice is a phase II trial and its population is not a population in organ failure. Of 222 patients, 106 had investigator-assessed visceral crisis, which is a clinical judgment rather than a physiological threshold3. The result argues strongly against reaching for combination chemotherapy in aggressive but compensated disease. It does not establish endocrine therapy with a CDK4/6 inhibitor as the answer in a patient with failing liver function. There the binding constraint is time to response, and median time to response was 4.9 months with the endocrine combination against 3.2 months with chemotherapy3.
Nobody has randomised a patient in organ failure. No trial named in this section carries time to symptomatic or biochemical improvement as a primary endpoint, which is the endpoint the decision needs. The operational question in true crisis is how fast a response is required. Where the answer is measured in weeks rather than months, the agent with the fastest response is right, and that is usually chemotherapy.
3 · Invasive lobular disease in the metastatic setting
Lobular carcinoma behaves differently in the metastatic setting, and most of the difference is a measurement problem rather than a treatment problem.
The pattern of spread is distinctive. Serosal surfaces, peritoneum, retroperitoneum, gastrointestinal tract and gynaecological organs are involved more often than in ductal disease. Bone involvement is common and is frequently diffuse rather than focal.
That pattern defeats the assessment tools. Lobular deposits infiltrate in single files without forming a mass, so they produce little to measure on computed tomography and can be invisible until an organ stops working. Linitis plastica of the stomach and diffuse peritoneal thickening are the classic traps. Response assessment by unidimensional criteria is therefore less reliable in this histology than in ductal disease, and a stable scan carries less reassurance.
The systemic treatment algorithm is the same. Lobular disease is overwhelmingly hormone receptor positive and HER2-negative, so it sits in this chapter by default. Two differences are worth holding. Loss of E-cadherin defines the histology and does not currently select a therapy. And the higher frequency of HER2-activating mutations in lobular disease makes sequencing worthwhile, which links this section to HER2-mutant disease as a distinct target.
4 · Long-term disease control and the cost of staying on treatment
Many patients now spend years on continuous treatment for this disease. That is the achievement of the last decade and it creates its own problem.
Three costs accumulate and none of them appear in a progression-free survival curve.
The first is physiological. Cumulative neuropathy, marrow suppression, ocular toxicity and metabolic effects build over lines. The distinction between cumulative and exposure-dependent toxicity, and what it implies for treatment breaks, is in Treatment-free intervals and cumulative toxicity.
The second is financial. Oral targeted agents given for years produce cost exposure that persists as long as the response does. This is the one cost that improves with a longer remission and worsens with it at the same time, and it is developed in Access, affordability, and financial toxicity.
The third is what a treatment schedule does to a life. Three-weekly infusions, monthly monitoring and the interval before each scan are the structure of the year. Patients rarely raise this unprompted.
In practice
Ask at every line what the treatment is for. Control of symptoms, control of a threatened organ, or extension of time are different aims, and only the first is verifiable within a month. Name the expected magnitude before starting. Medians in the pretreated setting are short, and a patient told that a line offers months rather than years is better placed to decide what to spend them on. Review the whole regimen annually rather than at progression only. Agents accumulate. Supportive medications outlive the episode that prompted them. Ask directly about cost. Patients who are rationing doses report it when asked and rarely otherwise, and a rationed dose will present as resistance Distinguishing resistance from under-dosing, non-adherence, and pseudo-progression.
References
All 4 references
Cardoso F, Paluch-Shimon S, Senkus E, et al. 5th ESO-ESMO international consensus guidelines for advanced breast cancer (ABC 5). Ann Oncol 2020 31:1623-1649. PMID 32979513
Cardoso F, Paluch-Shimon S, Schumacher-Wulf E, et al. 6th and 7th International consensus guidelines for the management of advanced breast cancer (ABC guidelines 6 and 7). Breast 2024 76:103756. PMID 38896983
Lu YS, Mahidin EIBM, Azim H, et al. Final results of RIGHT Choice: ribociclib plus endocrine therapy versus combination chemotherapy in premenopausal women with clinically aggressive hormone receptor-positive/human epidermal growth factor receptor 2-negative advanced breast cancer. J Clin Oncol 2024 42:2812-2821. PMID 38771995
Schmidt M, Thill M, Rey J, et al. First-line ET plus palbociclib versus standard mono-chemotherapy in high-risk HR-positive/HER2-negative metastatic breast cancer and indication for chemotherapy: primary results from the randomized phase IV PADMA study. ESMO Open 2026 11:107707. PMID 42275799