How to Write an Oncology Case Report

Oncology readers check stage, histology, biomarkers, response and toxicity against shared standards. A case report that skips those standards cannot be compared with anything.

Guide · 10 min read · Directive Publications Editorial Team

What makes an oncology case report different

An oncology case report follows the same basic framework as any case report: the CARE checklist, with its timeline, diagnostic assessment, intervention, outcomes and consent. What changes is how much of the content has a shared language.

Cancer care runs on standards. Tumours are classified by histology, staged with TNM, tested against biomarker panels, assessed for response with defined criteria, and their treatment toxicities graded on a common scale. Readers of a cancer case report check every one of those details, because they want to know whether your patient resembles theirs. CARE asks for "prognostic characteristics when applicable" in the diagnostic assessment; in oncology, that line carries the stage, grade and biomarker profile.

The cases worth writing up are usually one of these:

• An unusual response, such as a durable response to a drug in a tumour type where it is not expected to work
• An unexpected or poorly described toxicity, including immune-related adverse events
• A rare histology, site or presentation that changes how the diagnosis is made
• A diagnostic pitfall, such as a metastasis mistaken for a primary tumour
• A molecular finding that explains resistance or response

Whichever type you have, the steps below make it comparable with other patients.

Step 1: Pin down the histological diagnosis

Reviewers in oncology want to see how the diagnosis was reached, not just the name of the tumour.

• Specimen. Say whether the diagnosis came from cytology, a core biopsy, an excision or a resection, and from which site.
• Classification. Use the terminology of the current WHO Classification of Tumours for that organ system, and say which edition. Names change between editions, and a term from an older edition can make a case hard to find or compare.
• Grade and key features. Give the grade with the grading system used, and features such as margin status or lymphovascular invasion where they matter to the lesson.
• Immunohistochemistry. List the markers that decided the diagnosis, with positive and negative results. A rare-tumour report that says "immunohistochemistry confirmed the diagnosis" without naming the markers gives a reviewer nothing to check.
• Differential diagnosis. Name what was excluded and how.

Structured pathology datasets, such as the College of American Pathologists cancer protocols or the International Collaboration on Cancer Reporting (ICCR) datasets, are useful checklists for what a complete pathology description includes. If the pathologist who made the diagnosis made a substantial contribution, ask whether they meet your target journal's authorship criteria.

Step 2: Stage with a named system and version

Staging is where oncology case reports most often become ambiguous, because staging systems now change site by site.

The American Joint Committee on Cancer has moved from fixed editions to AJCC Version 9, released one disease site at a time. Cervix, appendix, anus, brain and spinal cord, neuroendocrine sites, vulva, lung, thymus, mesothelioma, nasopharynx and others now have Version 9 chapters, and AJCC states that all disease sites in the 8th Edition remain current until replaced. Separately, UICC published the 9th edition of the TNM Classification of Malignant Tumours in July 2025 and recommended it take effect from 1 January 2026.

So "stage IIIA lung adenocarcinoma" is incomplete. Write it out:

If the patient was staged under one version and you are writing after another took effect, report the stage as it was assigned and, if it is relevant to the lesson, note what it would be under the newer version.

Lymphomas, myeloma, leukaemias and central nervous system tumours use their own staging, risk or grading systems. Name the one you used and its version in the same way.

Step 3: Report molecular results precisely

Molecular findings are often the reason a cancer case report gets read, and also where precision slips most easily.

1. Name the test. Immunohistochemistry, FISH, PCR, or a sequencing panel, with the panel name or laboratory where appropriate.
2. Name the specimen. Tumour tissue, cytology, plasma (circulating tumour DNA) or a germline sample such as blood. Say whether the tissue was the primary tumour or a metastasis, and when it was taken relative to treatment.
3. Separate somatic from germline. A variant found in tumour tissue is not necessarily inherited. If germline testing was done, report it separately.
4. Use standard nomenclature. Write sequence variants in HGVS nomenclature, the internationally recognised standard maintained under the Human Genome Organization, with the reference transcript, and use approved HGNC gene symbols.
5. Give quantities with context. Report variant allele fraction, and for markers such as tumour mutational burden or PD-L1 expression give the assay, the scoring method and the threshold used.
6. Name any classification framework. If you call a variant actionable, say on what basis, for example the four-tier AMP/ASCO/CAP system for somatic variants published in 2017.

Step 4: Describe treatment and response with standard criteria

CARE asks for the type, dosage, strength and duration of each intervention and the reasons for any change. In oncology, write each treatment so another clinician could see exactly what the patient received:

• Systemic therapy by generic drug name, dose (for example per square metre of body surface area), schedule, cycle length and number of cycles
• Dose reductions, delays and discontinuations, each with its reason
• Radiotherapy as total dose, number of fractions and technique
• Surgery by procedure and resection status
• The line of therapy, and whether a multidisciplinary tumour board made the decision
• Off-label use, compassionate use or expanded access, stated plainly

For response, the RECIST working group publishes RECIST 1.1: up to five target lesions in total (two per organ), partial response as a decrease of at least 30% in the sum of diameters, and progressive disease as an increase of at least 20% with an absolute increase of at least 5 mm, or new lesions. If you use those terms, the measurements should support them. If the scans were not measured that way, describe what changed instead of borrowing the vocabulary.

Immunotherapy cases may need iRECIST, published by the same working group in 2017 because tumours can respond differently to immunotherapies than to chemotherapy. Lymphoma, myeloma and brain tumours have their own response criteria; name the ones you applied.

Imaging figures should show matched views: the same modality and comparable slices before and after treatment, with lesions marked and the time point labelled relative to treatment start.

Step 5: Grade toxicity and state follow-up honestly

Adverse events are reported in oncology using the Common Terminology Criteria for Adverse Events. NCI released CTCAE v6.0 in 2025, and v5.0 remains in use, including for ongoing NCI-sponsored studies. Report the version you used, the term, the grade, how the event was managed, and whether treatment was resumed, modified or stopped. For a toxicity case, also give the time from treatment start to onset and your reasoning about attribution.

Follow-up is where cancer case reports most often overreach. One patient's response does not show that a treatment works, and a short follow-up does not show durability. State:

• Follow-up duration from diagnosis and from the start of the treatment you are describing
• Disease status at last follow-up, in the same criteria you used for response
• Cause of death, if the patient died, and whether it was related to the cancer or its treatment

Prefer "no evidence of disease at 18 months" over "cured". The first is a finding; the second is a claim one case cannot support.

Step 6: Settle consent and identifiability

Consent for publication is covered in detail in our guide to patient consent for case reports and images. Cancer cases raise some specific issues:

• Rare tumours are identifiable. A rare histology plus age, sex, occupation and treating hospital can identify a patient even with the name removed. Leave out details that do not serve the lesson.
• Imaging carries hidden identifiers. Remove names, dates and hospital numbers burned into images and in file metadata. Head and neck imaging and three-dimensional reconstructions can show facial features.
• Genetic results affect relatives. A germline finding is also information about blood relatives. Discuss with the patient what will be published.
• The patient may have died. COPE guidance says consent should be required for any case report in which a person can be identified, including deceased people, and is then sought from the next of kin. Journal policies and local law vary, so check both.
• Trial patients. If the patient was treated in a clinical trial, check with the sponsor and your ethics committee before writing, and cite the trial registration.

CARE also asks for the patient's perspective where possible. A short account from the patient, or with consent from their family, is often what makes an oncology case memorable.

Choosing an oncology case report journal

There are three broad options. General case report journals take cases from every specialty. Oncology-specific case journals reach a cancer readership; for example, Case Reports in Oncology says in its current author guidelines that it requires a completed CARE checklist at submission and a statement of who gave written consent, and that a case report is typically 2,500 words with up to three tables and figures. Molecular case journals, such as Cold Spring Harbor Molecular Case Studies, focus on genomic and molecular analyses presented alongside the clinical picture, which suits a case whose lesson is the molecular finding.

Tumour-site and society journals sometimes have case or image sections too. If the case centres on a cancer operation, a surgical journal may ask for the SCARE guideline for surgical case reports rather than CARE. Check word limits, figure limits and fees before you draft, and use our guide to choosing where to publish a case report to build a shortlist.

Before you submit

1. Histological diagnosis uses current WHO terminology, with the edition named
2. Decisive immunohistochemistry markers listed with results
3. Staging system, version, prefixes, TNM categories and stage group all stated
4. Molecular tests named with specimen, somatic or germline status, and HGVS nomenclature
5. Every regimen given with drug names, doses, schedule, cycles and modifications
6. Response terms match the criteria actually applied
7. Adverse events graded with a named CTCAE version
8. Follow-up duration and status at last contact stated without overclaiming
9. Imaging and pathology figures stripped of identifiers
10. Consent on the journal's form, from the patient or next of kin
11. CARE checklist completed if the journal requests it

Directive Publications sets out its requirements in the author guidelines. When your case report is ready, you can submit your manuscript directly.