Part XI · Hormone receptor-positive, HER2-negative disease · Chapter 58

Biology and the treatment spine

Endocrine dependence is the thread. What establishes it, what sustains it, how it fails, and what is left when it has.

1 · The biology this part assumes

This chapter carries no figures. It is the spine of the part, and the chapters after it are the evidence. Where a claim here rests on a trial, the section that argues that trial is named.

The defining fact of this subtype is a dependency. The tumour uses oestrogen receptor signalling as a growth input, and removing that input is the single most effective thing oncology can do to it. Nothing else in breast cancer returns as much benefit for as little toxicity. The mechanism is set out in Estrogen receptor signaling and endocrine resistance, and the drugs that act on it in Endocrine therapy.

Three properties follow from the dependency, and each shapes a decision in this part.

It is slow. Proliferation is lower than in the other subtypes, so the disease takes longer to recur and longer to kill. That is why the recurrence curve in this subtype does not flatten at five years, and why the risk of a first distant recurrence is still real at fifteen. Dormancy, residual disease, and late recurrence is the chapter about that tail, and it is the reason extended endocrine therapy is a question at all.

It is measurable, but only as a continuum forced into a category. The receptor is scored by immunohistochemistry against a threshold, and the threshold was set by convention rather than by biology. A tumour at one per cent and a tumour at ninety-five per cent are both called positive and do not behave alike. Receptor assessment and the measurement problem is the measurement chapter, and most of the apparent disagreement between trials in this part is a disagreement about which populations their thresholds admitted.

It is losable. Endocrine dependence is a state rather than a fixed property, and the tumour can leave it. It can mutate the receptor so that the drug no longer fits, bypass the receptor through a parallel growth pathway, or abandon the lineage that made it dependent in the first place. Estrogen receptor signaling and endocrine resistance covers the first two and Plasticity and non-genetic heterogeneity the third. Which of these has happened is what the later-line chapters are trying to tell apart.

2 · What the subtype does to the treatment decision

Every decision in this part is a version of one question. How much treatment does this particular tumour need, given that the treatment works and that the patient will be taking it for years?

That framing is unusual in oncology and it is worth stating plainly. In the other two subtypes the dominant question is what to add. Here it is what to leave out, because the backbone already works and its costs accumulate over a decade rather than a course.

Four things follow.

The comparator is never nothing. A trial in this subtype adds to, or substitutes into, an endocrine backbone that the patient would have received anyway. A regimen that looks impressive against no treatment may add little to the backbone. This is why the control arm column is the first thing to read in every table in this part.

Risk is what allocates, and risk is estimated rather than measured. Clinicopathological risk, genomic assays and nodal status all estimate the same unobserved quantity, which is the probability that micrometastatic disease is already present. Prognostic and predictive assays and somatic profiling covers the assays. The chapters here treat a risk score as an eligibility criterion rather than as a fact about the patient.

Adherence is a treatment variable. An endocrine course runs five to ten years, and a substantial fraction of it is not taken. Every effect size in this part was achieved under real adherence rather than perfect adherence, which means the trial result already includes the loss. It cannot be claimed twice.

Time is the currency in advanced disease. The disease is usually incurable once it has spread and often measured in years. What a line of treatment buys is time before the next one, and time spent on a better-tolerated regimen. Deciding the order of treatments is therefore a real decision rather than a technicality, and General principles of metastatic management states the principles this part assumes.

3 · The spine, from diagnosis to last line

Read in order, the part answers six questions.

Does this early tumour need chemotherapy? The default is endocrine therapy. Chemotherapy is added where the risk is high enough that endocrine therapy alone is likely to be insufficient, and genomic assays exist to find the patients for whom it is not. Adjuvant chemotherapy and the genomic assay decision argues the assay trials and the populations they enrolled.

Which endocrine agent, and for how long? The choice follows the source of oestrogen, so it follows menopausal status. The duration question is separate from the agent question and has its own trial record. Adjuvant endocrine therapy carries both, including the extension trials and the fracture cost that comes with them.

Should a cell cycle inhibitor be added after surgery? Four trials asked this and two answered yes. The disagreement between them is not about the class, and Adjuvant CDK4/6 inhibition is the chapter that establishes what it is about instead.

Is there a reason to treat before surgery? In this subtype the neoadjuvant setting is a weaker instrument than it is in the others, because response is lower and pathological complete response carries less prognostic information. Neoadjuvant therapy says what the setting still buys, and The neoadjuvant setting as a scientific platform says what it is for in general.

What happens at first metastatic presentation? Endocrine therapy with a CDK4/6 inhibitor, for almost everyone, with the argument about who can safely be spared it. Endocrine resistance, and the presentations the sequence does not fit is the decision chapter and CDK4/6 inhibition, combinations and sequencing carries the trial record for the class.

What happens when that fails? The answer depends on how it failed. A mutated receptor is a reason to change the endocrine agent Endocrine therapy and receptor degraders after progression. An activated parallel pathway is a reason to add a pathway inhibitor, which Endocrine therapy and receptor degraders after progression also carries. A tumour that has left endocrine dependence altogether is a reason to stop trying to restore it, and Conjugates and chemotherapy in later lines is the chapter about what replaces it.

4 · Where this part disagrees with itself

Four places, and a reader who notices them is reading correctly.

The neoadjuvant and adjuvant chapters price response differently. In Neoadjuvant therapy a pathological complete response is a weak signal in this subtype. In Adjuvant chemotherapy and the genomic assay decision a response-guided strategy is treated as a live allocation tool. Both are defensible, and the difference is that the second is measuring a change in proliferation rather than a clearance of disease.

The duration evidence and the escalation evidence point in opposite directions. Adjuvant endocrine therapy shows that more endocrine therapy buys less than the first five years did. Adjuvant CDK4/6 inhibition adds two years of a second drug on top. The patient populations differ, and the chapters do not reconcile this because the trials do not.

First line and later line disagree about what a CDK4/6 inhibitor is for. CDK4/6 inhibition, combinations and sequencing treats the class as a first-line standard and also carries the evidence that continuing it past progression is weakly supported. The commonest escape mechanism does not need the kinase, which is the biological reason the two positions can both hold.

The advanced chapters disagree about how much a biomarker should decide. Endocrine therapy and receptor degraders after progression argues that an alteration should select the patient. Conjugates and chemotherapy in later lines argues that in later lines the prior-therapy history places the result more reliably than the biology does. Both are arguments about where the evidence is thin, not about what the drugs do.