Part XIII · Triple-negative disease · Chapter 75

Biology and the treatment spine

A subtype named for what it lacks, and the three things that turned out to be present after all.

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.

The defining fact of this subtype is a definition by exclusion. Three assays are negative, and the category is what remains. That is a statement about measurement rather than about biology, and almost everything difficult in this part follows from it. Triple-negative and other biologically diverse phenotypes is the chapter that takes the category apart, and Receptor assessment and the measurement problem is the one about the thresholds that draw its boundary.

Three consequences matter here.

The category is heterogeneous because nothing selected for homogeneity. Tumours with different transcriptional programmes, different immune infiltration and different repair capacity all satisfy the same three negatives. Trials enrol the category, so a positive trial is a statement about a mixture. Clinical and molecular classification carries the molecular classifications that cut it differently.

The disease is fast, and the timing of that speed is the clinical fact. Proliferation is high, recurrence is early, and the risk falls rather than persists. A patient who has not recurred by five years is in a different position from one with hormone receptor-positive disease at the same point. Dormancy, residual disease, and late recurrence is the chapter about why the two curves have different shapes.

Speed is also what makes the neoadjuvant setting work here. Response is common, it arrives quickly, and it carries more prognostic information in this subtype than in hormone receptor-positive disease. That is the structural reason the platform is used so heavily in this part, and The neoadjuvant setting as a scientific platform argues what the setting is for.

The three things that turned out to be present, after a category defined by absence, are the organising story of the part. Repair deficiency, in the germline and beyond it DNA repair, replication stress, and genomic instability. An immune microenvironment that can be engaged Breast cancer immunology. And surface antigens that a conjugate can reach, whether or not the tumour depends on them Antibody-drug conjugates and targeted delivery.

2 · What the subtype does to the treatment decision

Chemotherapy is the backbone and remains so. In the other two parts the backbone is targeted and chemotherapy is the thing being spared. Here it is the other way round, and every addition is an addition to cytotoxic treatment. That changes what a trial is testing and what its toxicity means.

The three positives are not alternatives. They are different questions asked of different patients. Germline status is inherited and binary. Immune infiltration is continuous, scored on an assay, and measured differently by different tests. Antigen expression is continuous and may not need to be high. Reading them as three branches of one algorithm is the commonest error in this part.

A biomarker threshold in this subtype is usually an eligibility criterion rather than a biological boundary. The clearest case is the assay used to select for checkpoint blockade, where the threshold moved and the populations of two trials therefore differ in a way that has nothing to do with the drug. The biomarker triage, and what is left when it finds nothing carries that argument and Receptor assessment and the measurement problem carries the measurement.

Early and advanced disease diverge more sharply here than elsewhere. In early disease the aim is cure and the evidence supports escalation. In advanced disease the disease is usually incurable and the record is a sequence of modest gains. The same drug class can be standard in one setting and marginal in the other, and this is not an inconsistency.

3 · The spine, from diagnosis to last line

Almost all early disease of any size is treated before surgery. The reason is the platform rather than the tumour: preoperative treatment turns an unmeasurable question into a measurable one, and what is found at surgery directs what follows. Selection by stage, and the price of treating everyone frames the early decision.

The preoperative regimen now includes checkpoint blockade for most patients, and what that settled is narrower than it is often read to be. Neoadjuvant therapy and checkpoint blockade is the chapter, and it separates what the evidence establishes from the four things it did not.

Residual disease after preoperative chemotherapy is the highest-yield finding in this part. It identifies a group whose risk is high and whose treatment changes because of it. Adjuvant and post-neoadjuvant therapy carries the post-neoadjuvant decision, and it is the place where germline status stops being an epidemiological fact and becomes a treatment allocation.

Advanced disease begins with the three tests. Germline status, the immune assay and antigen expression are established before the first line where possible, because each opens a different door and two of them close if the first line is chosen without them. The biomarker triage, and what is left when it finds nothing is the decision chapter and states the order in which to send them.

First line is chemoimmunotherapy for the assay-positive group and chemotherapy for the rest, with a PARP inhibitor as the alternative for germline carriers. Metastatic therapy from first line onwards carries the record from first line onwards.

After first line the part becomes a conjugate sequence. Two antigens, more than two drugs, and an unsettled question about what to do when one has already failed. Metastatic therapy from first line onwards argues it and Expression and alteration-defined treatment across subtypes carries the general form of the antigen-threshold problem it depends on.

4 · Where this part disagrees with itself

The early and advanced chapters price checkpoint blockade differently. In Neoadjuvant therapy and checkpoint blockade it is standard for most of the preoperative population regardless of the immune assay. In The biomarker triage, and what is left when it finds nothing and Metastatic therapy from first line onwards it is confined to the assay-positive group. The settings really do differ, and the honest statement is that the early result does not license the advanced use.

The category is used and criticised in the same part. Triple-negative and other biologically diverse phenotypes argues that triple-negative disease is not one thing. Every trial in Metastatic therapy from first line onwards enrolled it as though it were. The part does not resolve this, because the trials do not, and the resolution would require trials that have not been run.

Germline status changes the drug in some chapters and the risk in others. Adjuvant and post-neoadjuvant therapy treats it as an allocation criterion after residual disease. Inherited susceptibility and familial breast cancer treats it as an inherited risk with implications for the family. Both are correct and they are not the same use of the same fact.

Response means two different things across the part. In Neoadjuvant therapy and checkpoint blockade a pathological complete response is a prognostic endpoint strong enough to allocate further treatment. In Metastatic therapy from first line onwards a radiological response is a description of tumour shrinkage with a weaker relationship to how long the patient lives. The word is the same and the inferential weight is not.