The Pill That Could

Chapter 8: A Different Cancer

In a laboratory at the University of California, Los Angeles, in the autumn of 1986, an oncologist named Dennis Slamon is looking at a Southern blot.

It is a familiar piece of laboratory technology by now in this book — the dark bands on a gray film, the result of cutting DNA into fragments and asking which fragments contain a particular gene. The technique is the same one Annelies de Klein used in Rotterdam in 1982 to find ABL on the Philadelphia chromosome. Slamon is using it for a different question, in a different cancer.

He is asking how many copies of a particular gene a tumor has.

The gene is called HER2 — short for human epidermal growth factor receptor 2. The tumors are breast tumors, surgically removed from women with breast cancer, frozen, and shipped to UCLA from a network of hospitals around the country. Slamon has been collecting them for several years. He has, in his freezers, a tissue bank of several hundred breast cancer samples, each linked, with the patient's consent, to clinical follow-up information — how aggressive was the tumor, how big, how widely had it spread, did the patient relapse, did she survive.

He is looking, in the blot in front of him, at a row of lanes. Each lane is a different patient's tumor. Each lane shows the bands from a Southern blot probed with a small radioactive piece of HER2. In a normal cell, you would expect a single band of a particular intensity — the cell has two copies of HER2, one on each of its two copies of chromosome 17, and the band you see reflects those two copies. In some of the tumor lanes, the band is the expected intensity. In others, it is much darker. The cell, in those tumors, has more than two copies of HER2. Sometimes considerably more. In some of the lanes, the dark band is so much more intense than the normal band that the cell must contain dozens of copies of the gene.

Slamon counts the dark lanes. He is finding that they cluster, in his sample bank, with the worst-prognosis tumors.

Patients whose tumors had extra copies of HER2 — amplified HER2, in the language of the field — had relapsed sooner and died younger than patients whose tumors had normal HER2 levels. The pattern was strong enough, in his initial sample of about a hundred and ninety patients, that he could see it through the noise of the cohort. About a quarter of breast cancers had the amplification. Those breast cancers behaved differently from the others. They were faster, more lethal, and apparently a kind of breast cancer all their own.

Slamon, looking at the blot, is looking at the first piece of evidence that breast cancer is not one disease but several, and that one of those several is driven by a single specific molecular event in a way that, if you wanted to, you might be able to drug.

* * *

Breast cancer in 1986 was understood, by oncologists, as roughly one disease.

It was the most common cancer in American women. It killed about forty thousand of them every year. The treatment was surgery, supplemented by radiation, and — for the women whose tumors had been there long enough to spread to nearby lymph nodes, or who had the wrong kind of tumor on pathology — chemotherapy. The chemotherapy regimens were combinations of cytotoxic drugs adapted from the broader oncology pharmacopoeia: cyclophosphamide, methotrexate, fluorouracil; later, doxorubicin and the taxanes. They worked, in the sense that they extended life and increased the proportion of women cured by initial treatment. They did not, in most cases, work spectacularly well, and the side effects were the standard side effects of cytotoxic chemotherapy: hair loss, fatigue, nausea, low blood counts, some risk of long-term cardiac damage and second cancers.

There was, by the mid-1980s, one major exception to the cytotoxic paradigm. A drug called tamoxifen, originally developed in the 1960s as a possible contraceptive (it didn't work as a contraceptive; it had been a near-miss in a different program), turned out to block the estrogen receptor — a protein, found in the nucleus of certain cells, that responds to the female hormone estrogen by turning on a particular set of genes. About two-thirds of breast tumors expressed the estrogen receptor and depended, at least in part, on estrogen signaling for their growth. Tamoxifen, by blocking the receptor, slowed the growth of those tumors. It was given for years, sometimes decades, after surgery, as adjuvant therapy — the term oncologists use for treatment given after the primary tumor has been removed, to reduce the chance of recurrence.

Tamoxifen worked. It still works, in 2026, in essentially the same way, in essentially the same patients. It was, for the period in which it was the only such drug, the closest thing oncology had to a targeted therapy. It was relatively non-toxic compared to chemotherapy. It worked specifically on the tumors that had the right receptor. The tumors that didn't have the receptor were not affected by it, but they also weren't damaged by it, so it did no harm in those patients beyond not helping. The decision to give tamoxifen, beginning in the late 1970s, became one of the first applications of what would later be called precision medicine: you tested the tumor, and you treated based on what the test showed.

All 13 chapters of The Pill That Could · Library · All courses