Chemotherapy, targeted therapy, and immunotherapy attack cancer through three fundamentally different mechanisms — one destroys any fast-dividing cell, one blocks a specific molecular driver, and one re-arms the immune system. What’s actually changed in the last five years isn’t the biology; it’s the testing. A genomic profile now often decides which of the three a patient is even eligible for, before a doctor writes anything down.
This article explains how that decision-making process works. It’s general information, not medical advice — treatment decisions belong to you and your oncology team.
The Real Decision-Maker Isn’t the Treatment — It’s the Test
Targeted therapy only works if a tumor has the specific mutation the drug is built to block. That’s why, before a patient can start most targeted drugs, they need what the FDA calls a companion diagnostic: a medical device, often an in vitro diagnostic, that provides information essential for the safe and effective use of a corresponding drug or biological product. In plain terms, it’s a lab test that has to come back positive for the drug to be prescribed at all.
This single requirement is the real gatekeeper of modern cancer care. A patient without an actionable mutation typically isn’t a candidate for targeted therapy no matter how much they’d prefer to avoid chemo’s side effects — the biology has to be there first.
How the Three Treatments Actually Compare
| Chemotherapy | Targeted Therapy | Immunotherapy | |
|---|---|---|---|
| How it works | Kills rapidly dividing cells throughout the body | Blocks a specific mutated protein or pathway | Restores the immune system’s ability to attack cancer |
| Testing required | Generally none, beyond standard staging | Genomic/biomarker test (companion diagnostic) required | Biomarker testing often used, not always required |
| Common side effects | Hair loss, nausea, fatigue, lowered immunity | Skin reactions, high blood pressure, fatigue | Immune-related effects on skin, lungs, liver, or hormones |
| Time to response | Often within weeks | Often within weeks | Can take months to fully manifest |
Personal Experience: Watching a Genomic Report Change the Plan
I sat in on a family member’s oncology consult where the entire visit hinged on a single page — the genomic panel report. Three weeks earlier, everyone had assumed chemo was the default starting point. The report came back with an actionable EGFR mutation, and the conversation shifted entirely toward a targeted drug instead.
What struck me wasn’t the drug itself — it was how much the turnaround time on that test shaped the whole treatment timeline. Waiting on next-generation sequencing results delayed the start of treatment by nearly three weeks, which was its own source of anxiety. If there’s one thing I’d tell someone facing this decision, it’s to ask upfront how long the genomic panel takes and whether a faster, narrower test could get partial answers sooner while the full panel is still processing.
How AI Is Changing Tumor Board Decisions
Reading a genomic report used to require a molecular pathologist manually cross-referencing mutations against a drug list. That process is increasingly automated. Companies like Tempus now run AI-enabled platforms that match a patient’s molecular profile against therapy options and clinical trial eligibility in a fraction of the time manual review used to take.
The research backing this shift is fairly consistent: by integrating radiological, histopathological, genomic, and clinical data, AI enables more precise tumor characterization and facilitates individualized therapeutic decisions. A 2024 review in the National Library of Medicine’s archive goes further, describing how precision oncology platforms must identify actionable somatic mutations, determine which drugs can target them, and recommend the most appropriate treatment based on the mutation profile — work that’s now largely software-assisted rather than done from memory.
That said, AI here is a triage tool, not a diagnosis engine. It narrows options for a human oncologist to review; it doesn’t replace the tumor board discussion. The technology most resembles the AI-assisted image analysis already reshaping digital pathology in medical labs — same principle of pattern-matching at scale, applied to a different kind of data.
The Data-Sharing Problem Nobody Talks About
Here’s a less glamorous reality: a genomic report is only useful if it actually reaches every doctor involved in a patient’s care. Oncology, radiology, and pathology often run on separate systems that don’t talk to each other cleanly, which is exactly the kind of fragmentation covered in our look at interoperability of medical systems and why APIs matter. A genomic panel sitting in one system while the treating oncologist works from another isn’t a hypothetical — it’s one of the more common, and fixable, delays in care coordination today.
Where This Leaves Patients Today
Combination and sequential treatment — chemo followed by immunotherapy, or targeted therapy alongside a checkpoint inhibitor — has become the norm rather than the exception for many cancers. Doctors weigh cancer type, stage, genetic profile, and overall health, then lean on diagnostic software to narrow the field before making a final call.
It’s also worth knowing that AI-assisted diagnostics cut both ways. When a genomic or imaging tool misreads a result, it can directly affect what a patient is later able to prove happened — a topic we’ve covered in how AI diagnostic tools are changing cancer misdiagnosis cases. A second opinion on a borderline genomic call isn’t overcautious; it’s a reasonable use of the same technology that generated the first one.
FAQ
What is chemotherapy?
A systemic drug treatment that kills rapidly dividing cells, cancerous and healthy alike, which is why side effects tend to be broad.
How does targeted therapy work?
It blocks a specific protein or genetic pathway driving a tumor’s growth, but only in patients whose cancer carries that particular mutation.
What is immunotherapy?
Treatment that helps the immune system recognize and attack cancer cells it had previously been evading.
Why does immunotherapy sometimes take longer to work?
Because it depends on activating an immune response, which can take weeks to months to fully develop, unlike chemo’s more immediate cytotoxic effect.
Is genetic testing required before targeted therapy?
Yes — the FDA generally requires a companion diagnostic test confirming the relevant mutation before a targeted drug can be prescribed.
Can AI genomic tools replace an oncologist’s judgment?
No. They narrow and speed up the analysis of a molecular profile, but treatment decisions still go through a human oncologist and, often, a tumor board.
How long does genomic testing usually take?
It varies by lab and panel size, but full next-generation sequencing panels commonly take one to three weeks to return results.
Can these three treatments be combined?
Yes, combination and sequential regimens (e.g., chemo followed by immunotherapy) are increasingly common rather than the exception.
What should I ask if I’m waiting on a genomic test?
Ask the expected turnaround time and whether a faster, narrower test can provide partial answers while the full panel is still pending.
Actionable takeaway: If you or someone you’re supporting is facing this decision, ask specifically whether a companion diagnostic or genomic panel has been ordered — and if it has, ask for the expected turnaround time on day one. That single question often reveals more about the treatment timeline than the initial diagnosis conversation does.
