Oncology is the specialty that learned to fight cancer with smarter weapons, not bigger ones.
In April 2001, the New England Journal of Medicine published a trial of a new kind of cancer pill. One faulty enzyme, BCR-ABL, causes chronic myeloid leukaemia. The pill, later named imatinib, was built to block that enzyme. The trial gave it to 83 patients whose earlier treatment, interferon, had failed. Of the 54 who took 300 mg a day or more, 53 saw their blood counts return to normal, most within four weeks. The side effects were minimal. The drug had treated a cancer by aiming at the one molecule that drove it.
Eighteen months later, in October 2002, a much older trial gave its final answer. It had begun in 1976, when surgeons believed that removing more breast tissue meant more cure. The trial gave women with invasive breast cancer one of three operations: removing the whole breast, removing only the lump, or removing the lump and then giving radiation. After twenty years, 1,851 women had been followed. Survival did not differ between the three groups. Radiation after lumpectomy did matter for the breast itself: the cancer came back there in 14.3% of women, against 39.2% without it. Removing less had cost nothing in survival.
Eight years after that, in August 2010, the field tried a different weapon. Nothing had improved survival in melanoma that had spread. A trial gave 676 such patients, whose disease had grown despite treatment, a drug called ipilimumab. The drug does not attack the tumour. It releases a brake on the immune system’s T cells. Median survival was 10.0 months with ipilimumab and a vaccine, against 6.4 months with the vaccine alone. The drug had a real cost: severe immune side effects in 10 to 15% of patients. But the body’s own defences had been turned on a cancer, and people had lived longer.
These are three of the thousands of studies that shaped modern oncology. They taught oncology that a pill aimed at one molecule could control a leukaemia, that less surgery could cure as well as more, and that the immune system could be turned against a tumour, and each time the answer was in a paper years before it was in a textbook.
It is still happening. Recently, in November 2024, the New England Journal of Medicine published a trial in a form of metastatic bowel cancer that responds poorly to chemotherapy. The trial gave 303 patients who had not yet been treated either two immunotherapy drugs, nivolumab and ipilimumab, or chemotherapy. After two years, 72% on the immunotherapy were free of progression, against 14% on chemotherapy. The immunotherapy was also gentler: severe side effects in 23%, against 48%. For these patients, chemotherapy was no longer the first choice. Practice follows the literature as it goes. It is a good culture. It has exactly one cost.
The literature grows faster than anyone can read it.
Across the medical topics an oncologist needs to stay up to date on, from breast and colorectal cancer to immunotherapy and radiation oncology, around 1,900 full-text papers now arrive every week. That is about 8,200 a month, and more than 90,000 a year. Targeted therapy for cancer alone is 442 of each week’s 1,900. Breast cancer treatment is 179. These are not press releases or opinion pieces. They are the same kind of paper as the lumpectomy trial and the ipilimumab trial, and somewhere in each week’s 1,900 are the handful that will change a prescription.
It is tempting to believe that experience closes the gap, that twenty years of clinic teach what the papers teach. The evidence points the other way. A systematic review of 62 studies found that in most of them, more years in practice went with lower adherence to current standards of care. Experience is not what keeps an oncologist current. Reading is. And reading is the one thing the week does not have room for.
So we built the tool you need to stay up to date.
We summarise the latest research for you and send it directly to your email inbox, every week, or whenever you choose.
Two immunotherapy drugs held MSI-H colorectal cancer back far longer than chemotherapy
TakeawayIn patients with MSI-H or mismatch-repair-deficient metastatic colorectal cancer and no previous treatment for it, nivolumab plus ipilimumab left 72% free of progression at 24 months, against 14% with chemotherapy.
Patients whose metastatic colorectal cancer is MSI-H or mismatch-repair deficient do poorly on standard chemotherapy, with or without targeted drugs. CheckMate 8HW, a phase 3 open-label trial, randomised them to nivolumab plus ipilimumab, nivolumab alone, or chemotherapy.
- •Progression-free survival at 24 months: 72% (95% CI 64 to 79) vs 14% (95% CI 6 to 25) with chemotherapy (P < 0.001).
- •Restricted mean survival time at 24 months: 10.6 months longer (95% CI 8.4 to 12.9).