Before any treatment decision, a goal must be named. There are four, and they are not a progression toward one endpoint — they are genuinely different objectives. Cure means eliminating the cancer so the patient lives a normal lifespan. Control means slowing or halting progression when cure is no longer attainable. Palliation means relieving symptoms and improving quality of life without necessarily extending it. And comfort means that disease-modifying treatment no longer helps — the focus shifts entirely to the patient's remaining time. The goal usually shifts over a disease course, but matching treatment intensity to the current goal is the first clinical obligation. Aggressive multimodality therapy is right for a curable cancer. It can be the wrong choice — more harm than benefit — for a patient whose goal is comfort.

Cancer treatment uses three modalities. Surgery is the oldest and most effective for many localized solid tumors — clear margins, lymph-node sampling, minimal disruption to surrounding structures. Its limitation is reach: it removes what the surgeon can see, but cannot address microscopic dissemination beyond the operative field. Radiation uses ionizing radiation to damage DNA and kill cells — delivered as external beam, brachytherapy, or stereotactic approaches. It can be primary treatment, adjuvant after surgery, neoadjuvant before surgery, or concurrent with chemotherapy. Systemic therapy travels through the bloodstream and can reach disease anywhere in the body — chemotherapy, hormonal therapy, targeted agents, and immunotherapy. Each modality addresses what the others cannot. That is why they are combined.

Systemic therapy is not one thing — it is four distinct categories with different mechanisms and different evidence bases. Chemotherapy uses cytotoxic agents that disrupt DNA synthesis or the mitotic machinery, killing dividing cells. Hormonal therapy blocks hormones that drive cancer growth — estrogen for breast cancer, androgen deprivation for prostate. Targeted therapy uses small molecules or antibodies against specific molecular features of cancer cells: kinase inhibitors, HER2-directed agents. And immunotherapy activates the immune system against cancer — checkpoint inhibitors that release brakes on T cells, and CAR-T cell therapies that engineer immune cells to recognize tumor antigens. Each category arrived from a different scientific lineage and works by a different principle.

When systemic therapy or radiation is given relative to surgery is not a scheduling detail — it changes what the treatment is trying to accomplish. Neoadjuvant therapy is given before the primary intervention: it can shrink a breast tumor to enable lumpectomy where mastectomy would have been required, and reveals whether the cancer responds to a regimen before committing to the full course. Adjuvant therapy is given after surgery to a patient with no detectable disease — the rationale is that micrometastases are statistically likely to be present, and population-level trial evidence shows adjuvant therapy reduces recurrence. You are treating disease you cannot see. Concurrent therapy means two modalities simultaneously — chemoradiation for head and neck or cervical cancer. Consolidation deepens a response; maintenance prevents progression over the long term. Each term encodes a different mechanism and a different evidentiary basis.

The recurring failure mode in oncology evidence is treating the proxy as the thing itself. A drug that improves response rate and progression-free survival has demonstrated that it affects the tumor. It has not demonstrated that it benefits the patient until the hard endpoints move. The FDA's accelerated-approval pathway allows initial approval on surrogate endpoints, with confirmatory trials proving real benefit required afterward. Recent FDA initiatives — including Project Optimus, focusing on dose optimization for meaningful benefit — reflect awareness that approvals on surrogates have sometimes resulted in drugs that reached patients without ever demonstrating survival benefit. Some drugs have remained approved for years without completing confirmatory trials. The pattern is consistent: tumor shrinkage is what the drug does to the tumor. Overall survival and quality of life are what it does for the patient — and the two can diverge.

Because the decisions about how to combine surgery, radiation, and systemic therapy involve multiple specialties, treatment is planned by multidisciplinary tumor boards — surgeon, radiation oncologist, medical oncologist, pathologist, and radiologist together — rather than sequentially or siloed. The rationale for combination is that each modality addresses what the others cannot. Surgery removes the bulk primary tumor but leaves microscopic disease. Radiation sterilizes the local bed. Systemic therapy addresses distant micrometastases that neither surgery nor radiation can reach. No single specialist has the full picture. The tumor board is the mechanism by which the modalities are integrated into a coherent plan — and by which the treatment goal, and the endpoints that will measure its success, are agreed upon before treatment begins.

The chapter's claim is this: surrogate endpoints can move without hard endpoints following, so treatment claims must ultimately rest on overall survival and quality of life. A drug that improves response rate and progression-free survival has demonstrated that it affects the tumor. It has not demonstrated that it benefits the patient until the hard endpoints move. The finding that would revise this: rigorous, disease-specific evidence that a particular surrogate is a validated predictor of survival — a formal meta-analysis showing that in a given cancer, improvements in progression-free survival reliably and quantitatively translate into improvements in overall survival across many trials. Where such validation exists, the surrogate earns the right to stand in, and the skepticism relaxes. But that validation must be demonstrated, not assumed. Two questions remain genuinely open: for which cancers and which surrogates is progression-free survival actually a validated stand-in for overall survival — and when survival and quality of life diverge, how should the two hard endpoints be weighed against each other.