Skip to main content
Cancer Explained
Donate
Beginner 8 min readSource checked

Chemotherapy and Second Cancers

Why some cancer-fighting drugs can rarely cause a later cancer, which drugs are involved, and how care teams manage the risk

Source

National Toxicology Program — Report on Carcinogens, Fifteenth Edition

An older man, woman in a headscarf, and another woman talk together on a home sofa
An older man, woman in a headscarf, and another woman talk together on a home sofa

Key fact

The NTP Report on Carcinogens, Fifteenth Edition, lists cyclophosphamide, melphalan, chlorambucil, and thiotepa as known to be human carcinogens, each based on sufficient evidence from studies in humans.

The short answer

The NTP Report on Carcinogens lists cyclophosphamide, melphalan, chlorambucil, and thiotepa individually as known human carcinogens, each on human evidence. NCI's PDQ late effects summary splits therapy-related leukemia into an alkylating agent type with a 3-to-5-year latency and a topoisomerase II inhibitor type appearing 6 months to 3 years after exposure.

  • The NTP Report on Carcinogens, Fifteenth Edition, lists cyclophosphamide, melphalan, chlorambucil, and thiotepa as known to be human carcinogens, each based on sufficient evidence from studies in humans.

  • Every listing shows a dose relationship: leukemia risk rose with cyclophosphamide and melphalan dose, and rose with both dose and duration for chlorambucil.

  • NCI's PDQ late effects summary describes two therapy-related leukemias: an alkylating agent type with a latency of 3 to 5 years, and a topoisomerase II inhibitor type appearing 6 months to 3 years after exposure.

  • PDQ says risk of therapy-related leukemia plateaus after 10 to 15 years, and beyond 15 years the absolute excess risk is 0.02 cases per 1,000 person-years.

Choose how you want to understand this

The full explanation.

Why some cancer drugs are themselves carcinogens

The drugs that work by damaging DNA in cancer cells can damage DNA in healthy cells too. The US National Toxicology Program publishes a Report on Carcinogens. Its Fifteenth Edition names several of these drugs one by one, as known to be a human carcinogen.

That listing is not based on animal studies alone. For each of the drugs below, the RoC profile says the classification rests on sufficient evidence of carcinogenicity from studies in humans.

  • Cyclophosphamide was first listed in 1980. Its profile cites studies that found excesses of bladder cancer and leukemia in people treated with it.
  • Melphalan was first listed in 1980. Patients treated for breast cancer, ovarian cancer, or myeloma had a leukemia risk more than 100 times higher.
  • Chlorambucil was first listed in 1981. The cancer linked to it is mainly acute nonlymphocytic leukemia.
  • Thiotepa was first listed in 1981 on animal evidence, then reclassified as known to be a human carcinogen in 1998. Its human evidence is specifically leukemia.

None of that changes the arithmetic of treatment. It does mean the risk is specific and measured rather than vague.

Dose is not incidental to these findings

Every one of these listings carries a dose relationship, which is the strongest signal that the drug, and not something else, is doing the damage.

For cyclophosphamide, the RoC profile cites a German case-control study. In it, leukemia risk rose as the dose rose. It also cites a nested case-control study in non-Hodgkin lymphoma patients. There, bladder cancer risk rose as the total dose rose.

For melphalan, the profile says leukemia risk increased with increasing dose, and was not affected by co-exposure to radiation therapy.

For chlorambucil, the profile describes a randomized trial in 431 patients with polycythemia vera. Acute nonlymphocytic leukemia was 13 times more common with chlorambucil plus phlebotomy than with phlebotomy alone. Risk rose with both dose and duration.

Two different second leukemias, with two different clocks

NCI's PDQ summary on late effects of childhood cancer treatment separates therapy-related blood cancers into two types, and the difference is practical.

The alkylating agent type. PDQ names eight drugs here. They are cyclophosphamide, ifosfamide, mechlorethamine, melphalan, busulfan, the nitrosoureas, chlorambucil, and dacarbazine. Risk is dose dependent. Latency runs 3 to 5 years after exposure. It is tied to changes in chromosomes 5 and 7.

The topoisomerase II inhibitor type. PDQ names etoposide, teniposide, and drugs related to the anthracyclines. This type shows up as overt leukemia after 6 months to 3 years. It is tied to balanced translocations at chromosome bands 11q23 or 21q22.

PDQ adds that the names have changed. These are now called myelodysplastic syndrome post-cytotoxic therapy, and acute myeloid leukemia post-cytotoxic therapy. The older labels were therapy-related MDS and therapy-related AML.

When the risk peaks, and when it fades

PDQ gives a shape to the risk over time that is more reassuring than most people expect.

The short latency is under 10 years from the primary cancer diagnosis. Risk then plateaus after 10 to 15 years.

Beyond 15 years, PDQ says the risk of a later leukemia stays significantly raised. The standardized incidence ratio is 3.5. But PDQ puts that in absolute terms right away. These events are rare, with an excess risk of 0.02 cases per 1,000 person-years.

A large pooled analysis puts numbers on the chemo side. It drew on studies of children diagnosed between 1930 and 2000. In it, 147 cases of second primary leukemia were matched to 522 controls. Other treatments were controlled for. A topoisomerase II inhibitor dose above 2,500 mg/m², with no radiation, carried an odds ratio of 14.3 for second AML. The confidence interval was wide, from 2.7 to 75.1.

PDQ notes that second primary leukemias were most likely in the first decade after treatment.

Solid second cancers work differently

PDQ makes a distinction worth carrying into any survivorship visit.

Therapy-related solid second cancers make up 80% of all subsequent neoplasms. They track strongly with radiation exposure, not with chemo. Their latency exceeds 10 years. PDQ says the risk keeps rising with longer follow-up.

Three factors raise that risk: younger age at the time of radiation, a higher total radiation dose, and a longer period of follow-up afterward.

The clinical implication is direct. The drug question and the radiation question have different answers and different timelines.

The overall numbers, in context

The Childhood Cancer Survivor Study reported 30-year figures. Any subsequent neoplasm came to 20.5%. Nonmelanoma skin cancer made up 9.1%. Later cancers with malignant histology, leaving out that skin cancer, came to 7.9%. Meningioma came to 3.1%.

PDQ calls that a sixfold increased risk against the general population. It also defines the term. A subsequent neoplasm is a histologically distinct growth. It develops at least 2 months after treatment for the first cancer ends. And it may be benign or malignant.

PDQ notes these subsequent neoplasms are the leading cause of late mortality unrelated to relapse.

These figures come from childhood cancer survivors, followed for decades. They should not be read across to an adult treated once with a short course of chemotherapy.

What follow-up actually watches for

NCI defines late effects as problems that may not appear for months or years after treatment ends. They differ from side effects that ease once treatment stops.

Where the drugs above were used, four concrete questions matter. Which agents were given. What the total dose was, in mg/m². Whether radiation was part of the plan. And what the radiation field covered. Those four answers decide which timeline on this page applies.

When to get help sooner

A therapy-related leukemia or myelodysplastic syndrome shows itself through the blood, so the signs are bleeding, infection and breathlessness rather than a lump.

  • Call 911 or go to an emergency department if bleeding will not stop, you are coughing or vomiting blood, or breathlessness comes on suddenly. If you are still on chemotherapy, a temperature of 100.4°F (38°C) or higher belongs here too: CDC calls fever during chemotherapy a medical emergency.
  • Call your care team the same day if bruises appear where nothing hit you, small red dots spread over your skin, your gums bleed when you brush, or you keep picking up infections.
  • Call your care team within a day or two if tiredness has deepened without explanation, you are short of breath on stairs you used to manage, or night sweats and unexplained weight loss have set in.

The wider category of long-term monitoring is covered in survivorship. The blood cancer most often involved is described in leukemia. The difference between a classification and a personal probability is set out in hazard versus risk.

Sources

Words to know

Tap any term to see what it means.

Browse the full glossary →

Older woman with white hair sits on a yoga mat doing a seated twist in a sunny living room.

Common questions

Which chemotherapy drugs are actually classified as carcinogens?

The NTP Report on Carcinogens, Fifteenth Edition, lists several by name as known to be a human carcinogen, each on sufficient evidence from human studies: cyclophosphamide and melphalan, first listed in 1980, and chlorambucil and thiotepa, first listed in 1981. The cyclophosphamide profile cites excesses of urinary bladder cancer and leukemia; the melphalan profile reports a relative risk above 100 for leukemia in patients treated for breast, ovarian, or myeloma disease.

Does the dose matter?

Yes, and that dose relationship is a large part of the evidence. NTP cites a German case-control study where leukemia risk rose with cyclophosphamide dose, and a nested case-control study where bladder cancer risk rose with cumulative cyclophosphamide dose. For melphalan, leukemia risk rose with dose and was not affected by co-exposure to radiation. For chlorambucil, a randomized trial in 431 polycythemia vera patients found acute nonlymphocytic leukemia 13 times more common with chlorambucil plus phlebotomy than phlebotomy alone, rising with dose and duration.

How soon after treatment would a second leukemia appear?

It depends on the drug class. NCI's PDQ late effects summary says the alkylating agent type — cyclophosphamide, ifosfamide, mechlorethamine, melphalan, busulfan, nitrosoureas, chlorambucil, dacarbazine — has a latency of 3 to 5 years and involves chromosomes 5 and 7. The topoisomerase II inhibitor type — etoposide, teniposide, and anthracycline-related drugs — appears after 6 months to 3 years and involves translocations at 11q23 or 21q22.

Does the risk last forever?

PDQ says the risk of these therapy-related blood cancers plateaus after 10 to 15 years. Beyond 15 years the risk remains statistically elevated, with a standardized incidence ratio of 3.5, but PDQ states the absolute excess risk is 0.02 cases per 1,000 person-years. Solid second cancers behave differently: their latency exceeds 10 years and PDQ says risk continues to increase with longer follow-up, driven mainly by radiation exposure.

How common are second cancers overall?

The Childhood Cancer Survivor Study figures PDQ reports are 30-year cumulative incidence of 20.5% for any subsequent neoplasm, 9.1% for nonmelanoma skin cancer, 7.9% for subsequent cancers with malignant histology excluding that skin cancer, and 3.1% for meningioma — a sixfold increased risk versus the general population. These figures come from childhood cancer survivors followed for decades and should not be read across to an adult treated once with a short course.

Questions to ask your doctor

Being prepared helps you get the most out of your appointments. Save or print these questions.

Open my question list

Tap a question to save it to your list (kept on this device).

Human Connection Layer

Speak With Trained Specialists & Human Navigators

Cancer Explained provides educational guidance, but does not replace trained specialists, social workers, or your medical team.

Free & Confidential

Talk to a trained cancer information specialist

Free, confidential assistance from NCI Cancer Information Service via phone, chat, or email.

Contact your oncology team

Locate after-hours contact numbers, portal messages, or urgent triage phone lines.

Find a patient navigator

Get one-on-one help with appointments, logistics, translation, and care coordination.

Find a genetic counselor

Discuss inherited mutation risk, family history, and genetic testing options.

Find an oncology social worker

Access emotional counseling, family support groups, and mental health resources.

Find a financial navigator

Locate copay assistance foundations, grant programs, and lodging/travel support.

Find a clinical-trial specialist

Search matching studies and speak with NCI trial information specialists.

Get urgent help

Immediate emergency guidance for fever (>100.4°F during chemo), severe pain, or shortness of breath.

Help Us Improve This Guide

Did this explanation answer your question and help you determine your next step?

Know someone who needs this?

Plenty of people are looking for something like this and do not know where to start. If this would help a friend or someone you love, send it on — we have written an opening line so you do not have to stare at an empty message. You can change every word of it.

Email itText itWhatsApp

Your message is written and sent in your own email or messaging app — we never see who you send it to, and nothing is added to any list.

Knowledge Check

0 of 2 answered

  1. Q1.According to this article, how are people most often exposed to certain chemotherapy drugs?
  2. Q2.Certain chemotherapy drugs is most strongly linked to which cancer?

This self-assessment checks understanding of educational content only. It is not medical advice.

Prepared by Cancer Explained's AI-assisted editorial system

Written from federal health agency material and checked line by line against the source cited below.

Plain-language explanation of the published sources cited on this page. AI-assisted, source-checked, not clinician-reviewed.

Last updated: 2026-08-18Next planned review: 2028-07-05

How this page was created

Cancer Explained does not originate medical claims. Every page restates guidance already published by the National Cancer Institute, the CDC, the USPSTF and the FDA, in plain language, with the source cited so you can check the original yourself. AI does the translating and organizing; automated checks test claims, citations, clarity and safety before anything publishes. We do not employ clinicians and do not intend to — our work is translation and navigation, not clinical judgment. Nothing here is personal medical advice, and no page can account for your particular situation.

Editorial status — Source checked. This page was written with AI assistance and checked line by line against the sources listed on it. That confirms the sources support what the page says. It is not a medical review, and it does not confirm the page is complete or right for your situation.

General education — varies by person. Answers genuinely differ between people. This page explains what commonly varies and points you to your care team for your situation.

Human medical review: not completed. Pages here are not signed off by a clinician before they publish. That is not an oversight we are quietly working around: we restate published guidance and cite it, so the authority belongs to the source rather than to us, and every page names where its claims come from — you can verify us instead of trusting us. Where a volunteer clinician has reviewed a page, their name and credentials appear on it; where no name appears, no clinician has checked it. We are glad to have reviewers and are recruiting them, and we do not hold pages back waiting for one. Use this site to understand your situation and to ask better questions of the people treating you.

Our editorial processHow we use AIReport an error

How this page was created

Cancer Explained does not originate medical claims. Every page restates guidance already published by the National Cancer Institute, the CDC, the USPSTF and the FDA, in plain language, with the source cited so you can check the original yourself. AI does the translating and organizing; automated checks test claims, citations, clarity and safety before anything publishes. We do not employ clinicians and do not intend to — our work is translation and navigation, not clinical judgment. Nothing here is personal medical advice, and no page can account for your particular situation.

Editorial status: Source checked This page was written with AI assistance and checked line by line against the sources listed on it. That confirms the sources support what the page says. It is not a medical review, and it does not confirm the page is complete or right for your situation.

Human medical review: not completed. Pages here are not signed off by a clinician before they publish. That is not an oversight we are quietly working around: we restate published guidance and cite it, so the authority belongs to the source rather than to us, and every page names where its claims come from — you can verify us instead of trusting us. Where a volunteer clinician has reviewed a page, their name and credentials appear on it; where no name appears, no clinician has checked it. We are glad to have reviewers and are recruiting them, and we do not hold pages back waiting for one. Use this site to understand your situation and to ask better questions of the people treating you.

Read more about our editorial process, our use of AI, and our corrections policy.

Spotted a problem? Report an error — a factual mistake, broken or outdated source, confusing wording, or anything that seems unsafe. Please do not include names, medical record numbers, dates of birth, addresses, or other identifying medical information in your report.

After using this page, do you understand what to do next?

Anonymous — we only record the answer, never who gave it.