Part XVII · Research methods, evidence, and emerging therapy · Chapter 98

Unresolved questions and future directions

What this book cannot answer, stated as problems rather than aspirations.

1 · Preventing metastatic dissemination

Adjuvant therapy is built on a premise that the biology has already contradicted. The premise is that dissemination happens late, so removing the primary early and treating residual cells with systemic therapy should prevent metastasis.

Dissemination occurs early, often before a primary is clinically detectable, and this is set out in Metastatic dissemination and organ tropism. Cells leave during the earliest stages of lesion development. By the time a tumour is palpable or screen-detected, dissemination has in many cases already happened. The implication is uncomfortable. Earlier detection of the primary cannot prevent what has already occurred, which is one reason screening's mortality benefit is smaller than the shift in stage distribution suggests.

Preventing dissemination therefore requires acting on cells that have already left, or on the process that lets them survive after leaving. Neither is currently addressable. The steps are known individually: local invasion, intravasation, survival in circulation including the clustered form that is disproportionately efficient¹, extravasation, and survival at the distant site. What is not known is which step is rate-limiting in human disease, and a therapy aimed at a step that is not rate-limiting will not change outcomes.

The honest statement is that no current or near-term therapy targets dissemination. Everything in this book treats cells that have already disseminated or removes the tissue they might disseminate from.

2 · Eliminating dormant and residual disease

Late recurrence in hormone-receptor-positive disease is a distinct problem that adjuvant trial design is poorly shaped to address. The recurrence hazard continues for two decades, and half of recurrences occur after five years, which is the finding discussed in Endocrine therapy.

Three specific obstacles stand between the dormancy biology in Dormancy, residual disease, and late recurrence and a therapy.

Measurement. Dormant disseminated tumour cells cannot be counted with adequate sensitivity by any routine test. ctDNA detects proliferating disease, not quiescent cells. A therapy targeting a population that cannot be measured cannot demonstrate that it worked except through a twenty-year endpoint.

Mechanism. Dormancy is not one state. Cellular quiescence, immune-mediated equilibrium and angiogenic dormancy are different mechanisms with different vulnerabilities, and a tumour cell population may contain more than one. A therapy targeting a survival dependency of one state leaves the others.

Trial feasibility. The endpoint is recurrence at ten to twenty years in a population where most patients will not recur. This is the extreme case of the adjuvant design problem in Neoadjuvant, adjuvant, and metastatic design compared, and no current funding or regulatory structure accommodates it.

The realistic path is a validated intermediate endpoint for dormant burden. Without one, this problem does not become tractable.

3 · Overcoming heterogeneity and acquired resistance

Resistance is not a single failure mode and treating it as one has produced a decade of combination trials with disappointing results.

Pre-existing resistant subclones are present at diagnosis, below detection, and are selected by treatment^2,3. Nothing is acquired here except visibility. Acquired genetic alterations arise under treatment pressure, of which ESR1 mutation is the cleanest example^4,5. Non-genetic tolerance is a reversible transcriptional state that does not require mutation and is described in Plasticity and non-genetic heterogeneity. Microenvironmental protection is resistance conferred by the stroma or immune compartment rather than by the tumour cell.

Each of these requires a different response and only the second is addressed by sequencing a resistant tumour. A patient whose resistance is non-genetic will have no actionable finding on a resistance biopsy, and the negative result will be reported as absence of a target rather than as evidence about mechanism.

The unresolved question is whether resistance can be prevented rather than treated. Preventing it means acting before selection occurs, which means acting on a population that is not yet detectable, which returns to the measurement problem in Eliminating dormant and residual disease. SERENA-6 is the first evidence that acting at the moment of resistance emergence rather than at progression delays progression. Switching to camizestrant when an ESR1 mutation appeared gave median progression-free survival of 16.8 against 9.2 months, with a hazard ratio of 0.45. Second progression-free survival was 25.7 against 19.1 months, with a hazard ratio of 0.63⁶. Overall survival is not established, so what has been shown is a delay in progression rather than longer life. That is not prevention either, but it is the closest the field has come.

4 · Achieving durable immune control

Immunotherapy works in a minority of breast cancers and works durably in fewer. KEYNOTE-522 and KEYNOTE-355 established benefit in triple-negative disease^7,8. Hormone-receptor-positive disease, the majority, has shown little benefit.

The unresolved questions are specific.

Why is hormone-receptor-positive disease cold. Low mutational burden is part of it and is not sufficient, since some low-burden tumours in other diseases respond. Oestrogen signalling has immunomodulatory effects that are incompletely characterised. The answer determines whether the problem is addressable at all.

Which patients benefit. PD-L1 expression, tumour-infiltrating lymphocytes and mutational burden each predict imperfectly and none predicts well enough to allocate therapy confidently. The TIL work in Tumor-infiltrating lymphocytes, biology and standardized scoring is the strongest signal and is not sufficient as a single marker.

Why does benefit not persist. Some responses are durable and most are not. The mechanisms of acquired immune escape in Primary and acquired resistance to checkpoint blockade are described and not yet targetable.

What produces durable control. In the small number of patients with prolonged disease-free survival after immunotherapy, whether the tumour is eradicated or held in immune equilibrium is unknown. That distinction determines whether treatment can ever stop, which connects directly to Defining cure in metastatic disease.

5 · Defining cure in metastatic disease

A small number of patients with metastatic breast cancer live many years without progression. Whether any of them are cured is a question the field avoids because it cannot currently be answered.

The difficulty is definitional before it is empirical. Cure means no remaining malignant cells capable of producing disease. Available tools measure the absence of detectable disease, which is a different statement. ctDNA below the limit of detection is consistent with eradication and with a dormant population that sheds nothing, and Liquid biopsy and longitudinal monitoring cannot distinguish them.

The clinical consequence is concrete. A patient in prolonged complete response on maintenance therapy asks whether she can stop. There is no test that answers this. Stopping risks reactivation, and continuing carries cumulative toxicity and cost for years.

Three things would change this. An assay sensitive to residual and non-shedding disease. Evidence from prospective cessation trials in prolonged responders, which are difficult to run because patients and clinicians are reluctant to randomise. And a biological account of what sustains prolonged control, which would say whether it is eradication or equilibrium.

Until then the field should say what it knows. Some patients with metastatic disease live a very long time. Whether they are cured is not known, and describing them as cured is a claim the evidence does not support.

6 · Where heterogeneity sets a ceiling on precision medicine

This is the payoff of Part VI and the argument the book has been building toward. It is stronger than a caveat.

Precision medicine assumes that a measurement on a sample predicts the behaviour of the disease. Three findings, taken together, say that assumption fails in a specific and bounded way.

Spatial protein-level heterogeneity within single tumours produces co-existing receptor phenotypes that the tumour's assigned immunohistochemical classification does not resolve, and that heterogeneity rises with grade and is independent of genomic heterogeneity⁹. Temporal instability means the measurement was made on a tumour that no longer exists in that form, and receptor status changes between primary and metastasis¹⁰. And measurement itself is unreliable at the thresholds that now determine treatment, with pathologist concordance for HER2 at the low end of 25%¹¹ and computational models agreeing with each other at 65.1%¹².

The consequence is a ceiling, not a correction. If a biomarker measured on one core cannot represent the tumour, then a fraction of what is recorded as treatment failure is allocation failure, and no improvement in the drug addresses it. The conjugate cross-resistance problem in Antibody-drug conjugates and targeted delivery has the same shape: a patient may be assigned by a target measurement that does not describe the tumour being treated.

This does not mean precision medicine fails. It means its achievable accuracy is bounded by sampling and measurement rather than by biological knowledge, and that the binding constraint has moved. Raising the ceiling requires sampling the tumour differently, through imaging that surveys all disease¹³ or through ctDNA that samples all shedding lesions, rather than knowing more about targets.

7 · Aligning innovation with affordability and global access

The therapeutic advances in this book are unavailable to most women with breast cancer. Survival differs several-fold between high-income and low-income settings, and the gap is driven by stage at presentation and access to treatment rather than by biology^14,15.

The structural problem is that innovation and access are optimised by different systems. Drug development responds to markets that can pay. Essential medicines for breast cancer, including tamoxifen and basic cytotoxics, remain unavailable or unaffordable in many settings¹⁶, while conjugates costing many multiples of annual income per patient enter the same guidelines.

Financial toxicity is not confined to low-income countries. Patients in high-income systems experience material hardship, deplete savings and abandon treatment because of cost, and this is associated with worse outcomes^17,18. Cost is a clinical variable, and treating it as an administrative one is a category error.

Three directions have evidence behind them. Resource-stratified guidelines that specify what to do at each level of available resource rather than describing one standard most systems cannot meet¹⁹. De-escalation research, which reduces cost and toxicity simultaneously and is the one area where the interests of high and low resource settings align exactly. And biosimilar and generic availability, which has done more for global access to trastuzumab and endocrine therapy than any new agent.

The uncomfortable observation is that most of this book describes therapy that widens the gap it documents in Access, affordability, and financial toxicity. A field that measures progress only by what is newly possible will keep doing that.

8 · What a revised edition should contain

This section exists because the book is designed to be revised and because naming what is expected to change is more useful than pretending it will not.

Several questions should be answerable next edition. Whether ctDNA-guided intervention improves outcomes beyond the timing result in SERENA-6, with DARE and SURVIVE reading out. Whether conjugate sequencing after conjugate failure is effective, which practice already requires and evidence does not yet support. Whether the dormancy-directed approach in CLEVER produces a signal worth developing. Whether PROTAC degraders extend beyond the ESR1-mutant population where VERITAC-2 showed benefit²⁰. And whether computational HER2 scoring reaches the reproducibility that Computational pathology and imaging says it has not.

Several will not be settled and should be restated rather than quietly dropped. The surrogacy status of ctDNA. Whether dormant disease can be eliminated. Whether durable immune control is achievable in hormone-receptor-positive disease. Whether cure in metastatic disease can be defined operationally.

Two structural commitments should survive any revision. Numbers stay attached to their sources, and a claim that cannot be sourced is rewritten or removed rather than softened. And the distinction between biology and measurement is maintained in every chapter, because most apparent disagreement in this field is disagreement about assays being reported as disagreement about disease.

The chapters most likely to need rewriting are the ones in Part XIV. That is by design. A book organised so that the most volatile material is isolated and identified is a book that can be revised without being rebuilt.

References

All 20 references

1. Aceto N, Bardia A, Miyamoto DT, et al. Circulating tumor cell clusters are oligoclonal precursors of breast cancer metastasis. Cell 2014 158:1110-1122. PMID 25171411
2. Gundem G, Van Loo P, Kremeyer B, et al. The evolutionary history of lethal metastatic prostate cancer. Nature 2015 520:353-357. PMID 25830880
3. Yates LR, Knappskog S, Wedge D, et al. Genomic evolution of breast cancer metastasis and relapse. Cancer Cell 2017 32:169-184. PMID 28810143
4. Schiavon G, Hrebien S, Garcia-Murillas I, et al. Analysis of ESR1 mutation in circulating tumor DNA demonstrates evolution during therapy for metastatic breast cancer. Sci Transl Med 2015 7:313ra182. PMID 26560360
5. O'Leary B, Cutts RJ, Liu Y, et al. The genetic landscape and clonal evolution of breast cancer resistance to palbociclib plus fulvestrant in the PALOMA-3 trial. Cancer Discov 2018 8:1390-1403. PMID 30206110
6. Turner NC, Mayer EL, Park YH, et al. Switching to camizestrant at ESR1 mutation emergence before disease progression during first-line treatment of hormone receptor-positive advanced breast cancer (SERENA-6). Lancet Oncol 2026. PMID 42442380
7. Schmid P, Cortes J, Pusztai L, et al. Pembrolizumab for early triple-negative breast cancer. N Engl J Med 2020 382:810-821. PMID 32101663
8. Cortes J, Rugo HS, Cescon DW, et al. Pembrolizumab plus chemotherapy in advanced triple-negative breast cancer. N Engl J Med 2022 387:217-226. PMID 35857659
9. Mardamshina M, Karagach S, Mohan V, et al. Integrated spatial proteomic analysis of breast cancer heterogeneity unravels cancer cell phenotypic plasticity. Nat Commun 2025 16:10482. PMID 41290667
10. Priedigkeit N, Hartmaier RJ, Chen Y, et al. Intrinsic subtype switching and acquired ERBB2/HER2 amplifications and mutations in breast cancer brain metastases. JAMA Oncol 2017 3:666-671. PMID 27926948
11. Robbins CJ, Fernandez AI, Han G, et al. Multi-institutional assessment of pathologist scoring HER2 immunohistochemistry. Mod Pathol 2023 36:100032. PMID 36788069
12. McKelvey B, Torres-Saavedra PA, Li J, et al. Agreement across 10 artificial intelligence models in assessing HER2 expression in breast cancer whole-slide images. Mod Pathol 2026 39:100944. PMID 41489584
13. Parihar AS, et al. 18 F-Fluoroestradiol PET/CT for predicting benefit from endocrine therapy in patients with ER-positive breast cancer. J Nucl Med 2025. PMID 40081952
14. Allemani C, Matsuda T, Di Carlo V, et al. Global surveillance of trends in cancer survival 2000-14 (CONCORD-3): analysis of individual records for 37 513 025 patients diagnosed with one of 18 cancers from 322 population-based registries in 71 countries. Lancet 2018 391:1023-1075. PMID 29395269
15. Anderson BO, Ilbawi AM, Fidarova E, et al. The Global Breast Cancer Initiative: a strategic collaboration to strengthen health care for non-communicable diseases. Lancet Oncol 2021 22:578-581. PMID 33691141
16. Fundytus A, Sengar M, Lombe D, et al. Access to cancer medicines deemed essential by oncologists in 82 countries: an international, cross-sectional survey. Lancet Oncol 2021 22:1367-1377. PMID 34560006
17. Ramsey SD, Bansal A, Fedorenko CR, et al. Financial insolvency as a risk factor for early mortality among patients with cancer. J Clin Oncol 2016 34:980-986. PMID 26811521
18. de Souza JA, Yap BJ, Wroblewski K, et al. Measuring financial toxicity as a clinically relevant patient-reported outcome: the validation of the COmprehensive Score for financial Toxicity (COST). Cancer 2017 123:476-484. PMID 27716900
19. Anderson BO, Shyyan R, Eniu A, et al. Breast cancer in limited-resource countries: an overview of the Breast Health Global Initiative 2005 guidelines. Breast J 2006 12 Suppl 1:S3-S15. PMID 16430397
20. Campone M, De Laurentiis M, Jhaveri K, et al. Vepdegestrant, a PROTAC estrogen receptor degrader, in advanced breast cancer. N Engl J Med 2025 393:556-568. PMID 40454645