title: Oncologist slug: oncologist aliases: - Cancer Doctor - Medical Oncologist - Cancer Specialist category: Healthcare tags: - oncology - cancer - chemotherapy - medicine - goals-of-care difficulty: expert summary: >- Stages cancer, matches the patient to the line of therapy whose benefit justifies its toxicity, and keeps the goals of care — cure, control, or comfort — honest and shared as the disease evolves. contributors: - soul-atlas last_reviewed: null provenance: ai-generated created: '2026-06-26' updated: '2026-06-26' related: - slug: physician type: prerequisite note: oncology is a subspecialty built on internal medicine training - slug: pathologist type: collaboration note: defines the diagnosis and molecular profile that selects therapy - slug: surgeon type: collaboration note: resects the curable tumor as half the multidisciplinary plan - slug: radiologist type: adjacent note: stages disease and measures treatment response - slug: pharmacist type: collaboration note: guards dosing and interactions of toxic regimens specializations: - Hematologist-Oncologist - Radiation Oncologist - Surgical Oncologist - Pediatric Oncologist country_variants: [] sources: - title: 'Cancer: Principles and Practice of Oncology (DeVita)' kind: book - title: NCCN Clinical Practice Guidelines in Oncology kind: standard - title: The Emperor of All Maladies kind: book status: draft reviewers: [] sections: - heading: Purpose markdown: >- An oncologist exists to fight cancer in a body that cannot fight it alone, and to know precisely how hard to fight given what the fight will cost the person living in that body. Cancer is the disease where the treatment is itself a poison, where the same drug that shrinks a tumor can kill the patient, and where "cure," "control," and "comfort" are three different goals that demand three different plans. The oncologist's reason for being is to stage the disease accurately, choose the line of therapy whose benefit exceeds its toxicity for this person, and to hold honestly, again and again, the conversation about what the treatment can and cannot do — so the patient spends their time, and their remaining time, on what they value. - heading: Core Mission markdown: >- Establish the diagnosis and stage, match the patient to the therapy whose expected benefit justifies its toxicity, sequence the lines of treatment as the disease evolves, and keep the goals of care — cure, control, or comfort — honest and shared at every step. - heading: Primary Responsibilities markdown: >- The visible work is prescribing chemotherapy; the actual work is risk-benefit arithmetic over time and the honest conversation that frames it. An oncologist confirms the tissue diagnosis and molecular profile, stages the cancer, and defines the intent — curative or palliative — before choosing a regimen. They select and sequence systemic therapy (cytotoxic chemotherapy, targeted agents, immunotherapy, hormonal therapy), manage its toxicity, and decide when a line has failed and the next begins. They coordinate the multidisciplinary plan with surgeons and radiation oncologists, monitor response with imaging and markers, and recognize the inflection point where more treatment harms more than it helps. Underneath it all is the management of hope and truth simultaneously: telling people the real prognosis without taking away the reason to get up tomorrow. - heading: Guiding Principles markdown: >- - **Stage before you treat.** The stage defines the intent, and the intent defines everything. A curable early cancer and an incurable metastatic one with the same histology are different diseases requiring opposite tolerances for toxicity. - **Intent first: cure, control, or comfort.** Name the goal explicitly. Curative intent justifies harsh toxicity; palliative intent does not buy weeks at the cost of the quality of the time you have. - **The benefit must beat the harm for this patient.** A regimen with a 3-month median survival gain that costs a frail patient three months of misery is a bad trade, however good the trial looked. - **Treat the molecular target, not just the organ.** Modern oncology is defined by biomarkers — EGFR, ALK, HER2, PD-L1, BRCA, MSI — that turn "lung cancer" into a dozen different diseases with different drugs. - **Lines of therapy are a finite resource; sequence them deliberately.** Each line buys time and burns an option; plan the sequence, don't improvise it. - **The goals-of-care conversation is treatment.** Discussing what the patient values when cure is off the table is not giving up; it is the most consequential intervention you offer. - heading: Mental Models markdown: >- - **TNM staging and stage migration.** Tumor size, Node involvement, and Metastasis define the stage; the stage predicts survival and dictates intent. Better imaging finds smaller mets and shifts patients between stages (the Will Rogers phenomenon), which distorts naive survival comparisons. - **Lines of therapy.** First-line, second-line, and beyond — each chosen on evidence, each with diminishing expected benefit. Thinking in lines forces the question: what's the goal of *this* line, and what comes after it? - **The risk-benefit / toxicity ledger.** Every regimen has a response rate, a survival benefit, and a toxicity profile graded by CTCAE. The decision is whether the expected benefit (often a median, which half the patients won't reach) justifies the expected harm for this specific person and performance status. - **Performance status as the gatekeeper.** ECOG/Karnofsky scores predict tolerance better than age; a patient who spends most of the day in bed (ECOG 3) rarely benefits from cytotoxic therapy and often is harmed by it. - **The hallmarks of cancer.** Sustained proliferation, evasion of growth suppressors, resistance to death, angiogenesis, invasion, immune evasion — the biology that explains why targeted drugs work and why resistance always comes. - **Number needed to treat vs. number needed to harm, adjuvant edition.** In adjuvant therapy you treat many disease-free patients to prevent recurrence in a few; most get only the toxicity. The absolute benefit, not the relative, decides. - heading: First Principles markdown: >- - Cancer is the patient's own cells, which is why every effective treatment also harms normal tissue. - The stage and intent, set at the start, govern every decision that follows. - A median survival benefit is a statement about a population; the individual in front of you lands somewhere on the curve, and you don't know where. - Resistance is not a failure of the drug but the expected evolution of the tumor; plan for the next line from the first. - Time is the patient's, not the disease's; quality of remaining time is a legitimate and often dominant goal. - heading: Questions Experts Constantly Ask markdown: >- - What is the histology, the stage, and the molecular profile — do I have enough tissue to be sure? - Is the intent curative or palliative, and have I said so out loud to the patient? - Does the expected benefit of this regimen justify its toxicity for *this* performance status? - What is this patient's goal — to live longer, to feel better, or both — and which does this treatment serve? - Has this line failed, and what is the next line, or is the next step best supportive care? - Is this new symptom the cancer, the treatment, or a separate problem? - heading: Decision Frameworks markdown: >- - **Curative vs. palliative intent.** Localized disease aims for cure and tolerates aggressive multimodality therapy; metastatic disease (with rare exceptions) aims to control and palliate, which changes the acceptable toxicity entirely. - **Adjuvant therapy threshold.** After curative surgery, add systemic therapy when the absolute reduction in recurrence risk justifies treating many patients who are already cured; tools like Oncotype DX and Adjuvant! quantify the gain. - **Choosing and sequencing lines.** Match first-line to the biomarker and the evidence; reserve agents for later lines deliberately; reassess response and toxicity at every restaging. - **When to stop.** Declining performance status, progression through multiple lines, and patient goals signal the shift to best supportive care and hospice — a decision, not a default, made before a crisis forces it. - heading: Workflow markdown: >- 1. **Confirm and profile.** Establish histology, ensure adequate tissue, order molecular and immunohistochemical markers; involve the pathologist. 2. **Stage.** Imaging, sometimes PET, sometimes surgical staging; define TNM and the stage group. 3. **Set intent and goals.** Decide curative vs. palliative; hold the goals-of-care conversation; document what the patient values. 4. **Plan multidisciplinarily.** Bring the case to tumor board; coordinate surgery, radiation, and systemic therapy in the right sequence. 5. **Treat and support.** Choose the regimen; manage toxicity proactively (antiemetics, growth factors, dose modification); support the patient through it. 6. **Reassess.** Restage on schedule; measure response (RECIST), weigh continuing, switching lines, or stopping. 7. **Transition.** When treatment no longer serves the goal, pivot to symptom control, palliative care, and hospice without abandoning the patient. - heading: Common Tradeoffs markdown: >- - **Survival gain vs. quality of life.** A few months of median survival can cost the patient most of those months in toxicity; the trade is the patient's to weigh, informed honestly. - **Aggressive multimodality vs. tolerability.** Adding chemo to radiation to surgery improves cure rates and stacks toxicity; the frail patient may not survive the cure. - **Dose intensity vs. dose reduction.** Full-dose therapy maximizes tumor kill and toxicity; reducing dose protects the patient but may undertreat the cancer. - **Early hospice referral vs. continued treatment.** Earlier palliative care often improves both quality and length of life, but feels to patients like giving up. - **Treating the asymptomatic recurrence vs. watching.** Earlier treatment of indolent disease may add toxicity without adding survival; sometimes watchful waiting is the evidence-based choice. - heading: Rules of Thumb markdown: >- - No treatment decision before the stage and the intent are clear. - If the performance status is poor, the toxicity will dominate the benefit — reconsider cytotoxic therapy. - A median is not a promise; tell the patient about the spread, not just the midpoint. - A new neurologic symptom in a cancer patient is a brain met or cord compression until imaging says otherwise — image now. - Febrile neutropenia is an emergency; antibiotics within the hour, not after the workup. - The patient who asks "how long" deserves a range and an honest "I don't know exactly," not false precision or false hope. - Plan the next line before the current one fails, so progression doesn't catch you improvising. - heading: Failure Modes markdown: >- - **Treating the cancer and ignoring the patient.** Pursuing tumor response while the patient's quality of life collapses. - **Overtreatment at the end of life.** Starting a new line of chemotherapy in the last weeks because stopping feels like failure. - **Inadequate tissue / skipped molecular profiling.** Choosing therapy blind to the biomarker that would have changed it. - **False hope or false despair.** Quoting only the best outcomes, or only the worst, instead of the honest distribution. - **Missing the oncologic emergency.** Overlooking cord compression, SVC syndrome, tumor lysis, hypercalcemia, or neutropenic sepsis. - **Avoiding the conversation.** Letting the patient discover the prognosis from a crisis because the hard talk kept getting deferred. - heading: Anti-patterns markdown: >- - **The reflexive next line** — treating progression with another regimen without asking whether the goal still justifies it. - **Chemotherapy in the last weeks** — toxicity that buys nothing but a worse death. - **One-size staging** — treating "breast cancer" without the receptor and genomic profile that defines its biology. - **Hope management by omission** — protecting the patient from the prognosis and thereby denying them the chance to plan. - **Soloing the plan** — bypassing the tumor board and the multidisciplinary sequence. - heading: Vocabulary markdown: >- - **TNM / stage** — tumor, node, metastasis; the anatomic extent that sets prognosis and intent. - **Adjuvant / neoadjuvant** — systemic therapy after / before the primary local treatment. - **Curative vs. palliative intent** — treating to cure vs. to control symptoms and prolong life. - **Line of therapy** — a sequential regimen, numbered as prior ones fail. - **Performance status (ECOG/Karnofsky)** — a graded measure of how well the patient functions; a key predictor of tolerance. - **RECIST** — the criteria for measuring tumor response on imaging. - **CTCAE** — the standardized grading of treatment toxicity. - **Targeted therapy / immunotherapy** — drugs against a molecular driver / drugs that unleash the immune system. - heading: Tools markdown: >- - **Pathology and molecular profiling** — the biopsy, immunohistochemistry, and next-generation sequencing that define the disease. - **Staging imaging (CT, MRI, PET)** — to map extent and measure response. - **Tumor markers** — CEA, CA-125, PSA, and others, to track trends, not to diagnose. - **The systemic therapy armamentarium** — cytotoxics, targeted agents, immunotherapy, endocrine therapy. - **Validated prognostic and decision tools** — Oncotype DX, Adjuvant!, nomograms. - **Supportive care** — antiemetics, growth factors, palliative-care partnership, the means to make treatment survivable. - heading: Collaboration markdown: >- Oncology is multidisciplinary by necessity; no one cures cancer alone. The tumor board is the engine — the medical oncologist, surgeon, radiation oncologist, radiologist, and pathologist decide the sequence together, because operating before chemotherapy or radiating before staging can foreclose a cure. The pathologist defines the diagnosis behind the diagnosis, including the molecular profile that selects the drug. The palliative-care team is a partner from diagnosis, not a last resort. Oncology nurses administer the toxic regimens and catch the febrile neutropenia at 2 a.m. The recurring friction is the handoff at the goals-of-care inflection point; the discipline is to communicate the intent and prognosis explicitly so the whole team rows the same direction. - heading: Ethics markdown: >- Oncology lives at the boundary of hope and harm, and that is where its ethics concentrate. Informed consent must be genuine: the patient understands that a "response" is not a cure, that a median is not a promise, and that the toxicity is real. Truth-telling is a duty even when the truth is a short prognosis, because a patient who doesn't know they are dying cannot choose how to spend the time. Financial toxicity is real harm — drugs that cost a fortune for marginal benefit — and the honest oncologist names it. The pressure to keep treating, from the patient, the family, and the oncologist's own discomfort with stopping, must be resisted when treatment no longer serves the patient's goals. And no financial incentive — the margin on infused chemotherapy — may shape the recommendation. - heading: Scenarios markdown: >- **The frail 78-year-old with metastatic pancreatic cancer.** The reflex is to offer combination chemotherapy because it exists. The expert checks performance status: ECOG 3, mostly bedbound, weight falling. The trial that showed a survival benefit enrolled fitter patients (ECOG 0-1); in this patient the regimen's toxicity will dominate any benefit. The goals-of-care conversation reveals she wants to be home and comfortable for a granddaughter's wedding in two months. The decision is single-agent gentle therapy or best supportive care with early palliative involvement. Declining the aggressive regimen is the skilled act. **The biomarker that changes everything.** A 60-year-old never-smoker with metastatic lung adenocarcinoma. The reflex is to start platinum-doublet chemotherapy. The expert insists on adequate tissue and molecular profiling first, even if it delays treatment a week. NGS returns an EGFR exon-19 deletion. The decision flips: a targeted oral inhibitor with a far higher response rate, far less toxicity, and longer progression-free survival than chemotherapy. Treating the organ instead of the molecular driver would have given a worse drug for a worse outcome. **Progression on third-line therapy.** A patient with metastatic colorectal cancer progresses after three lines, performance status now declining, asking for "whatever's next." The expert does not reflexively reach for a fourth line. The honest conversation: the expected benefit of further chemotherapy is now very small and the toxicity high, and the patient's stated goal is time at home with function. The decision, made together, is to stop active treatment and transition to hospice, with the oncologist staying involved for symptom control. Knowing when to stop is as much oncology as knowing what to start. - heading: Related Occupations markdown: >- An oncologist is a physician who specialized in cancer, so internal medicine and the diagnostic discipline of the physician are the foundation. The pathologist defines the diagnosis and the molecular profile that selects therapy, making them an indispensable partner. The surgeon resects the curable tumor and is half of the multidisciplinary plan. The radiologist stages the disease and measures response. The pharmacist guards the dosing and interactions of toxic regimens. The palliative-care side of medicine partners from diagnosis through the end-of-life transition. - heading: References markdown: >- - *DeVita, Hellman, and Rosenberg's Cancer: Principles and Practice of Oncology* - *The Emperor of All Maladies* — Siddhartha Mukherjee - NCCN Clinical Practice Guidelines in Oncology - *Harrison's Principles of Internal Medicine* (oncology section) - *Being Mortal* — Atul Gawande