Oncology clinical trials differ from trials in nearly every other area of medicine in fundamental ways, from who participates to how success is measured to how drugs get approved. Cancer’s severity and the toxicity of its treatments force researchers, regulators, and patients into a different set of rules. If you’re considering joining a cancer trial, have a loved one in one, or are simply curious about how cancer drug development works, these differences matter.
Phase 1 Trials Start With Patients, Not Healthy Volunteers
In most areas of medicine, the earliest human trials recruit healthy volunteers. The goal is straightforward: give the drug to people with no underlying illness and see how the body handles it. Cancer trials flip this entirely. Because cancer drugs are designed to kill cells or disrupt powerful biological pathways, exposing a healthy person to them could cause serious, lasting harm. Cytotoxic chemotherapy, for example, attacks rapidly dividing cells without distinguishing between cancer and normal tissue. Giving that to someone without cancer would mean all risk and zero benefit.
Instead, oncology Phase 1 trials enroll patients who already have advanced cancer, typically people who have exhausted standard treatment options. This changes the entire calculus of early testing. Participants aren’t just helping generate safety data. They’re hoping, even with low odds, that the experimental drug might help them. It also means the earliest dose-finding data comes from people whose bodies are already affected by disease and prior treatment, which adds complexity to interpreting results.
How Doses Are Chosen Is Changing
For decades, oncology dose-finding followed a simple, aggressive logic: escalate the dose until side effects become intolerable, then back off slightly. That ceiling is called the maximum tolerated dose, or MTD. This made sense for traditional chemotherapy, which has a steep relationship between dose and effect. More drug killed more cancer cells, and patients and doctors accepted brutal side effects because alternatives were scarce.
Modern targeted therapies and immunotherapies don’t work this way. A kinase inhibitor or antibody designed to block a specific molecular pathway often reaches its full effect well below the dose that causes serious toxicity. Pushing to the MTD can saddle patients with unnecessary side effects that erode quality of life and force them to stop treatment early, ultimately reducing the drug’s benefit. The FDA has formally acknowledged this problem, noting that the MTD paradigm “can result in a recommended dosage that may be unnecessarily high, poorly tolerated, and adversely impacts functioning and quality of life.”
The shift now is toward finding an “optimized dosage,” one that maximizes benefit relative to harm by considering a wider range of data: low-grade side effects, how the drug behaves in the body, and how tumor activity responds at different dose levels. This is a significant departure from the traditional oncology playbook and one that trials in other fields never had to make, since they weren’t starting from an assumption that severe toxicity was acceptable.
Success Is Measured Differently
The gold standard endpoint in cancer trials is overall survival: how long patients live. But waiting for survival data can take years, especially as treatments improve and patients live longer. This creates a practical problem. If a promising drug takes five additional years to prove it extends life, patients who could benefit right now are waiting.
To address this, oncology trials rely heavily on surrogate endpoints, measurements that can be assessed sooner and are thought to predict longer-term benefit. The most common are progression-free survival (how long before the cancer starts growing again) and objective response rate (the percentage of patients whose tumors shrink by a defined amount). These let researchers gauge a drug’s activity in months rather than years.
Not all surrogates are equally reliable, though. Recent analysis in small-cell lung cancer found that progression-free survival correlates strongly with overall survival in the first-line treatment setting, making it a dependable stand-in. Objective response rate, however, showed no meaningful correlation with overall survival in either first- or second-line treatment. A drug that shrinks tumors doesn’t necessarily help patients live longer. This distinction matters because regulatory decisions and treatment changes sometimes hinge on these shorter-term numbers. Trials in cardiology or diabetes also use surrogates (blood pressure, blood sugar), but the relationship between those markers and long-term outcomes is generally better established.
Drugs Can Reach Patients Before Full Proof Exists
The FDA’s Accelerated Approval Program exists specifically for drugs that treat serious conditions with unmet needs, and oncology dominates its use. Under this pathway, a drug can be approved based on a surrogate endpoint, such as tumor shrinkage rate, before the lengthy process of confirming that it actually extends life.
This is a regulatory bargain. Patients get access to potentially effective treatments faster, but the drug company must then run confirmatory trials proving real clinical benefit. If those follow-up studies fail to show the drug works as hoped, the FDA can pull it from the market. This has happened with several cancer drugs in recent years. In most other therapeutic areas, drugs go through the full traditional approval process before reaching patients, making accelerated approval a defining feature of the oncology landscape.
Placebos Are Used Sparingly and Carefully
In many types of clinical trials, comparing a new drug to a placebo (an inactive pill) is standard practice. Conditions with high placebo response rates, fluctuating severity, or spontaneous remission often require a placebo arm to prove a drug genuinely works. Cancer is different. You generally cannot withhold effective treatment from someone with a life-threatening disease and give them a sugar pill instead.
When placebos do appear in cancer trials, they come with strict ethical guardrails. They’re most commonly used when no effective standard treatment exists for that specific cancer, or they’re added on top of standard treatment (so one group gets standard care plus the new drug, while the other gets standard care plus a placebo). The placebo in those cases helps maintain blinding so that neither patients nor doctors unconsciously bias their assessment of side effects and outcomes.
Even then, ethical review requires that patients assigned to placebo are not substantially more likely to die, suffer irreversible harm, or experience serious discomfort compared to those receiving the experimental treatment. Full disclosure of the trial design to participants is mandatory. This careful approach to control groups makes oncology trial design more complex and sometimes makes it harder to generate the clean comparative data that regulators want.
Biomarkers Shape Who Can Enroll
Increasingly, cancer trials require patients to have a specific genetic mutation or molecular marker in their tumor before they can participate. This is a direct consequence of precision medicine: if a drug targets a particular protein that only some cancers produce, testing it in an unselected population would dilute the signal and waste time. Biomarker-driven trial design is far more common in oncology than in most other fields.
This precision comes with a recruitment problem. When a trial requires a mutation found in only 3% or 5% of patients with a given cancer type, finding enough eligible participants becomes a persistent challenge. Analysis of biomarker-driven cancer trials found that low or no accrual was a problem across all trial phases. Overall, low accrual was the most common reason cancer trials stopped (28.7% of failures), and for biomarker-specific trials, the difficulty of finding eligible patients was a central driver of failure.
The screening process itself adds time and cost. Patients need tumor biopsies or blood tests analyzed with specialized molecular profiling before they can even be considered. Many patients who want to participate are screened out. This bottleneck doesn’t typically exist in trials for conditions like hypertension or depression, where eligibility criteria are based on symptoms or standard lab values rather than rare molecular features.
Costs Are Exceptionally High
Cancer clinical trials are among the most expensive in medicine. An analysis of federally sponsored cancer trials found median drug costs alone of $3.6 million for Phase 1 trials, $9.4 million for Phase 2, and $38.8 million for Phase 3. The mean costs were dramatically higher, with Phase 3 trials averaging $244.9 million in drug costs, reflecting a small number of extremely expensive studies that pull the average up. In comparative Phase 2 and 3 trials, experimental drug arms cost a median of $21.8 million while control arms cost $1.4 million.
These figures cover drug costs only, not the imaging, lab work, staff time, data management, and patient support that make up the rest of a trial’s budget. The high costs reflect several oncology-specific factors: expensive biological drugs, long treatment durations, intensive monitoring with frequent imaging, the complexity of biomarker screening, and the need for large trials to detect survival differences that may amount to weeks or months. Combined with the recruitment challenges of biomarker-driven designs, the financial and logistical barriers in oncology trial development significantly exceed those in most other therapeutic areas.