A first-year oncology trainee is asked why chemotherapy causes hair loss, mouth sores, and low blood counts. She gives the textbook answer: chemotherapy kills rapidly dividing cells. Cancer cells divide fast, so they are hit hardest. Her attending nods, then asks: why does high-dose chemotherapy wipe out hematopoietic stem cells, which divide faster than many slow-growing tumor cells? If faster division were the whole story, the marrow would survive and the tumor would die. It is often the reverse.

Cancer chemotherapy began with a wartime accident. In 1943, an explosion in Bari harbor released nitrogen mustard into the water. Soldiers exposed developed profound suppression of bone marrow and lymphoid tissue. The inference was immediate: a chemical that depletes proliferating cells might deplete a proliferating tumor. By 1946, nitrogen mustard had been used to treat lymphoma at Yale, and cytotoxic chemotherapy was born. The premise was productive — it generated decades of drug development and genuinely curative regimens. It is also incomplete in a specific and important way.

Many cancer cells do not divide faster than normal cells. Hematopoietic stem cells in the marrow divide more rapidly than many tumor cells. Chemotherapy works as well as it does not merely because tumor cells divide fast, but because cancer cells carry specific defects — in DNA repair, in apoptosis regulation, in metabolic flexibility — that make them less able to survive the damage the drug inflicts. A normal stem cell takes a DNA hit, activates its checkpoint, repairs the break, and continues. A tumor cell with defective mismatch repair or mutated p53 takes the same hit and dies. The anti-tumor effect tracks vulnerability, not speed.

The cell cycle runs G1, S, G2, and M. Cells not actively cycling sit in G0. Antimetabolites jam the DNA synthesis machinery, so they are S-phase-specific: they work only when the cell is actively copying its DNA. A cancer cell in G0 is unharmed; the same cell cycling into S phase two days later is vulnerable. S-phase-specific drugs therefore benefit from prolonged exposure to catch cells as they enter the vulnerable window. Alkylating agents, by contrast, can damage DNA in any phase, including resting cells. This phase-independence is one reason alkylating agents at high doses can ablate hematopoietic stem cells that an antimetabolite would miss.

Five drug classes, five distinct mechanisms. Alkylating agents form covalent crosslinks in DNA — cyclophosphamide requires mesna to prevent hemorrhagic cystitis; cisplatin cures most metastatic testicular cancers. Antimetabolites mimic normal building blocks to jam DNA synthesis — they are S-phase specific. Doxorubicin intercalates DNA and generates free radicals but causes cumulative cardiomyopathy capped at roughly 500 milligrams per square meter; bleomycin causes pulmonary fibrosis with a similar lifetime ceiling. Microtubule inhibitors — vincas and taxanes — arrest cells in mitosis and cause peripheral neuropathy. Etoposide, a topoisomerase II inhibitor, carries a one to two percent risk of secondary leukemia from induced chromosomal translocations.

Three regimens show the principle in practice. MOPP — mechlorethamine, vincristine, procarbazine, prednisone — was the first regimen to cure what had been uniformly fatal Hodgkin lymphoma. Different mechanisms, largely non-overlapping toxicities. R-CHOP adds rituximab, a monoclonal antibody targeting CD20, to an alkylator, an anthracycline, a microtubule inhibitor, and a steroid — the standard of care for diffuse large B-cell lymphoma. BEP — bleomycin, etoposide, cisplatin — cures most metastatic testicular cancer. A disseminated solid tumor, curable. The curative potential of BEP was among the most striking demonstrations of combination chemotherapy's power.

So here is the chapter's claim. Chemotherapy's anti-tumor effect is driven primarily by tumor-specific vulnerabilities in DNA repair, apoptosis, and metabolic flexibility — not merely by faster proliferation. The slogan tracks the toxicities well enough that it feels complete. It cannot explain why slow-growing lymphomas respond, why some fast-dividing tissues recover, or why combination regimens are built the way they are. Resistance arises when those defects are restored. What remains open: why predictive biomarker-driven selection has lagged so far behind mechanistic understanding, and whether the best cytotoxic combinations have already been found or simply displaced by attention to targeted therapy.