How Cancer Drugs Work: The Cell Cycle, Log-Kill and Why We Combine Them

Cancer is a disease of division

Strip cancer down to one sentence: cells that proliferate without a brake. Normal tissue divides on demand and stops on cue. A cancer cell has broken both signals — it grows when it shouldn't and ignores the orders to die (apoptosis). Because the defining behaviour is uncontrolled division, the oldest and still-central drug strategy is to attack cells while they are dividing. That is why classical chemotherapy hits DNA and the machinery of cell division — and, unavoidably, why it also hits the normal tissues that divide fastest: bone marrow, gut lining, hair follicles. The strategy is powerful precisely because it is not fully selective.

The cell cycle: four phases, and where drugs strike

A dividing cell moves through an ordered loop. The cell cycle has four working phases: G1 (the cell grows and prepares), S (it copies its DNA), G2 (it checks the copy and readies for division), and M (mitosis — it splits into two). Cells not currently cycling rest in a quiet state called G0. This map matters clinically because different drug classes ambush the cell at different points. Some drugs only work if the cell is in a particular phase; others do their damage no matter what the cell is doing. That single distinction — cell-cycle-specific versus cell-cycle-non-specific — organizes the entire cytotoxic pharmacy.

Cell-cycle-specific (CCS) drugs need the cell to be actively cycling, and usually in one phase. Antimetabolites — methotrexate, 5-fluorouracil, cytarabine — mimic the building blocks of DNA and jam replication, so they strike in S phase. The vinca alkaloids (vincristine, vinblastine) and the taxanes (paclitaxel, docetaxel) attack the mitotic spindle, so they strike in M phase. Because they need dividing cells, CCS drugs are most effective against rapidly proliferating tumours and are often given by infusion over time, to catch cells as they enter the vulnerable phase.

Cell-cycle-non-specific (CCNS) drugs do not wait for a phase. Alkylating agents (cyclophosphamide, cisplatin — the platinums behave similarly) crosslink and damage DNA directly; anthracyclines (doxorubicin) intercalate DNA and poison topoisomerase II. Because their damage does not depend on the cell being in S or M, they can injure even slow-dividing and resting (G0) cells — which makes them valuable against bulky, slow-growing tumours where much of the mass is not actively cycling. Their dose-response tends to rise with the size of the dose, which is why they anchor high-dose regimens.

The log-kill hypothesis: a fraction, not a number

This is the idea from the scene, made formal. The log-kill hypothesis, from the classic experiments of Skipper and Schabel, states that a given dose of a drug kills a constant FRACTION of the tumour cells present — not a constant number. If a dose achieves "3-log kill," it destroys 99.9% of the cells whether there are a billion or a thousand. The consequence is profound: because you are always killing a percentage of what remains, you can never subtract your way to zero in a single blow. Ten billion cells, knocked down by 99.9%, still leaves ten million. That is why treatment comes in repeated cycles, each one carving the same fraction off a smaller and smaller remnant.

Two practical rules fall straight out of log-kill. First, cure requires driving the burden very low — ideally to zero, but realistically low enough that the body's own immune surveillance can mop up the last few cells; that takes multiple cycles, continued past the point of visible disease. Second, timing matters: cycles are spaced to let normal tissues (marrow especially) recover, but not so far apart that the tumour regrows its losses between doses. The entire schedule of chemotherapy is a negotiation between killing the maximum fraction of tumour and giving the patient's healthy dividing cells time to recover.

Growth fraction: why fast leukaemias yield and slow tumours resist

The growth fraction is the proportion of cells in a tumour that are actively dividing at any moment. It is the hidden variable behind why some cancers melt away and others barely flinch. A high growth-fraction tumour — many acute leukaemias, some lymphomas — has most of its cells cycling, so cell-cycle-specific drugs find plenty of targets and responses can be dramatic. A bulky solid tumour, by contrast, often has a low growth fraction: its core is poorly perfused and many cells sit quietly in G0, out of reach of phase-specific drugs. Paradoxically, shrinking a big tumour can raise its growth fraction — surviving cells re-enter the cycle — which is one rationale for giving cell-cycle-specific drugs after debulking.

Why we combine drugs

A single agent almost never cures a cancer. Combinations do. Combination chemotherapy follows a few disciplined rules. Choose drugs each active against the tumour on its own. Choose different mechanisms of action, so they attack the cell in independent ways. Choose non-overlapping toxicities, so each can be given at full dose without stacking the same side effect. And, where possible, hit different phases of the cell cycle, so cells escaping one drug are caught by another. Above all, combinations delay resistance: the chance that a single cell carries resistance to three unrelated drugs at once is vanishingly small compared with resistance to one.

Drug resistance: how tumours fight back

Resistance is the reason a tumour that shrank at first can return unmoved by the same drug. It arises because tumours are genetically unstable and mutate constantly, so a resistant clone can be selected and expand. The mechanisms are worth knowing by name. Efflux pumps: the MDR1 gene makes P-glycoprotein, a membrane pump that ejects many structurally unrelated drugs (anthracyclines, vinca alkaloids, taxanes) — a single change conferring multidrug resistance. Altered or amplified targets: methotrexate resistance from amplified dihydrofolate reductase; a mutated target the drug no longer fits. Enhanced DNA repair: cells that repair alkylator and platinum damage faster survive it. Apoptosis evasion: a cell with a broken death programme (e.g. loss of p53) absorbs the damage without dying. Decreased activation or increased inactivation of the drug rounds out the list.

From poison to precision: cytotoxic → targeted → immune

Everything above describes classical cytotoxic chemotherapy. The field did not stop there. Cytotoxic drugs attack division itself, and pay for it in toxicity to every dividing tissue. The next generation, targeted therapy, aims at a molecular lesion the cancer depends on — imatinib blocking the BCR-ABL kinase in chronic myeloid leukaemia, trastuzumab against HER2-amplified breast cancer — sparing much of the collateral damage. The newest wave, immunotherapy, does not attack the tumour directly at all; it releases the brakes on the patient's own immune system — the checkpoint inhibitors against PD-1 and CTLA-4 — so the body's T cells destroy the cancer. These do not replace the principles above so much as build on them; combinations, resistance, and dosing logic all reappear in new forms. The chapters ahead in this Oncology section take each family in turn.