The selectivity problem
When we treat a bacterial infection, antibiotics exploit selective toxicity: bacteria have cell walls and ribosomes different enough from ours that a drug can poison them while sparing us. Cancer offers no such gift. A cancer cell is your own cell that has acquired mutations and lost its brakes on division. Its proteins, its DNA, its metabolism are almost identical to those of your healthy tissues. There is no obvious foreign target waiting to be hit.
So the earliest anticancer drugs aimed at the one thing cancer cells do more than most healthy cells: they divide, constantly. A cytotoxic drug poisons the machinery of cell division. Cancer cells, dividing fast, get hit hard. But so do your fast-dividing healthy tissues — bone marrow, gut lining, hair follicles. This is the central bargain, and the central tragedy, of classical chemotherapy.
Kill fraction and log-cell-kill
Chemotherapy does not kill a fixed number of cancer cells per dose — it kills a fixed fraction. This is the log-cell-kill model. If one treatment cycle kills 99.9% of tumor cells, it does so whether the tumor has a billion cells or a thousand. That sounds encouraging, but it has a hard consequence: you can never quite reach zero by killing fractions. A tumor of 10¹² cells reduced by 3 logs per cycle still leaves survivors after each round.
Log-cell-kill in action (1 log = 90% killed) Starting tumor burden: 10^12 cells (~1 kg, symptomatic) Cycle 1 (kills 3 logs): -> 10^9 cells Cycle 2 (kills 3 logs): -> 10^6 cells Cycle 3 (kills 3 logs): -> 10^3 cells (undetectable on scans!) Cycle 4 (kills 3 logs): -> 10^0 = 1 cell Lesson: scans go 'clear' around 10^9, but cure needs MORE cycles after that, because a fraction (not a number) dies each time.
This is why courses run for multiple cycles spaced apart. The gaps let your normal bone marrow recover between hits, since healthy stem cells bounce back faster than the tumor regrows. The whole mechanism of curative chemotherapy is a race: knock the tumor down by logs, rest, repeat, and try to reach the last cell before resistance or toxicity wins.
Cell-cycle thinking
To understand the drug families ahead, hold one picture in mind: the cell cycle. A cell copies its DNA in S phase, divides in M phase (mitosis), and rests or prepares in the G phases. Some antineoplastic drugs only harm a cell during a specific phase — these show cell-cycle specificity. Others damage cells in any phase, including resting cells. This single distinction explains how the drugs are scheduled and combined.