Oncology Clinic Notes

⚠ ECOG = WHO = Zubrod — the SAME 0–4 scale, three names (guidelines write "ECOG/WHO PS"; if a colleague says "WHO score 3" they mean ECOG 3). Karnofsky (KPS) is the separate 0–100 scale (10-point steps; 100 = normal). Performance status is THE gate for treatment fitness — it decides systemic therapy vs best supportive care, often more than the tumour itself.

Rule of thumb: 0–1 (KPS ≥70) → fit for systemic therapy; 2 (KPS 50–60) → borderline / attenuate; 3–4 (KPS ≤40) → usually best supportive care.

⚠ Treating a curable metastatic cancer palliatively is one of the gravest errors in oncology. Some cancers are curable even when widely metastatic, and they must be treated with full-dose curative intent: GERM-CELL / testicular (>80% cured even poor-risk, and still curable at relapse — see §6) · Hodgkin lymphoma · aggressive lymphomas (DLBCL) · gestational trophoblastic neoplasia (curable even with widespread mets) · most paediatric cancers.
WHY the difference: these tumours are intrinsically, exquisitely chemosensitive — germ-cell tumours largely retain wild-type p53 and undergo apoptosis readily after DNA damage, so there is no resistant reservoir to select for. Most epithelial solid tumours behave the opposite way.
WHY metastatic epithelial cancer is usually NOT curable: ① burden — a 1 cm lesion is ~10⁹ cells, and undetectable micrometastases are already seeded everywhere; ② clonal heterogeneity — treatment applies Darwinian selection, and resistant subclones pre-exist in proportion to burden (Goldie-Coldman); ③ DORMANCY — disseminated cells sit quiescent in niches (classically bone marrow in breast cancer) for years-to-decades, and chemo kills dividing cells, so quiescent cells survive — this is why breast cancer relapses at 10–20 years; ④ sanctuary sites (CNS/blood-brain barrier); ⑤ immune evasion already selected for.
⚠ BUT "incurable" is softening at the edges: OLIGOMETASTATIC disease treated with local therapy (metastasectomy/SBRT) produces genuine long-term survivors — resected colorectal liver mets, bone-only breast, sarcoma lung mets, oligomet RCC; and immunotherapy now yields a durable plateau in melanoma/NSCLC that looks like functional cure in a minority. The practical model for metastatic HR+ breast or CML-like disease is CHRONIC CONTROL over many years, not imminent death.

A known benign liver or lung lesion must NOT be read as a metastasis on the staging scan — that error turns a curable cancer into a false "stage IV" and changes the whole plan. Common benign mimics: liver — haemangioma, focal nodular hyperplasia, hepatocellular adenoma (HCA), simple cysts, biliary hamartomas; lung — granuloma, hamartoma. Rule: flag any pre-existing lesion to radiology explicitly as known, long-standing, stable and compare against old imaging; if it is solitary/equivocal and its interpretation would change management, get tissue or dedicated characterisation (MRI/contrast) rather than upstage on assumption. ⚠ HCA is oestrogen-driven (grows with OCP/pregnancy; rupture risk if >5 cm) — so in an HR+ breast cancer it also argues against tamoxifen (a partial hepatic oestrogen-agonist) and toward an aromatase inhibitor.

PREVENT, don't rescue — schedule the antiemetics BEFORE chemo, not PRN. Two things drive the choice: the regimen's emetogenic RISK (table) and the PHASE you're covering — ACUTE (0–24 h, serotonin/5-HT3-driven → 5-HT3 + steroid) · DELAYED (days 2–4/5, substance-P/NK1-driven → the NK1 + steroid ± olanzapine phase; classic with cisplatin, AC, carboplatin) · ANTICIPATORY (a LEARNED reflex before the next cycle → benzodiazepine, but the real fix is good control from CYCLE 1) · BREAKTHROUGH (nausea despite prophylaxis → add a DIFFERENT class, given ROUND-THE-CLOCK not PRN).

THE AGENTS + typical adult doses (local order-set governs — confirm your unit's protocol):
• 5-HT3 (serotonin) antagonist — ondansetron 8–16 mg IV / 8 mg PO · granisetron 1 mg IV (or transdermal patch) · palonosetron 0.25 mg IV (long half-life → also helps DELAYED). Best for ACUTE. SE: constipation, headache, QT prolongation.
• NK1 (neurokinin-1) antagonist — aprepitant 125 mg PO day 1 → 80 mg PO days 2–3 · fosaprepitant 150 mg IV day 1 (single dose covers 3 days) · NEPA (netupitant + palonosetron) = one oral dose. Covers DELAYED. ⚠ aprepitant inhibits CYP3A4 → reduce the dexamethasone dose when co-given; also interacts with warfarin.
• Dexamethasone — 12 mg IV/PO day 1 (range 8–20), then 8 mg PO daily days 2–4 for delayed. Broad benefit, acute + delayed. SE: hyperglycaemia, insomnia, gastritis, mood.
• Olanzapine (atypical antipsychotic, blocks multiple receptors) — 5–10 mg PO at night, days 1–4 (5 mg = similar efficacy, less sedation). Excellent for DELAYED + BREAKTHROUGH. SE: sedation (give nocte), increased appetite.
• Dopamine antagonists — the BREAKTHROUGH / prokinetic drugs — metoclopramide 10–20 mg PO/IV every 6–8 h · prochlorperazine 10 mg · haloperidol 1–2 mg. ⚠ EXTRAPYRAMIDAL reactions (acute dystonia, akathisia) — especially in the YOUNG — keep courses short.
• Benzodiazepine — lorazepam 0.5–1 mg the night before + morning of, for ANTICIPATORY nausea.

HOW IT ACTUALLY RUNS — IV in the chair, then pills for home:
① ACUTE — in the infusion unit, ~30 min BEFORE chemo (mostly IV): the 5-HT3 (IV) + dexamethasone (IV) + the NK1 (fosaprepitant IV, or aprepitant PO) ± olanzapine. ← "the steroid and the injections before chemo."
② DELAYED — take-home ORAL, days 2–4: dexamethasone PO + aprepitant PO (days 2–3 if the oral form was used) + olanzapine PO. ← "the pills for home."
③ BREAKTHROUGH — PRN at home if it breaks through: metoclopramide (± prochlorperazine / olanzapine / lorazepam) — and once it breaks through, give it round-the-clock, not only when nauseated.
🔴 THE COMMON FAILURE — seen with P1 (18M on cisplatin, 23 Jul): the take-home breakthrough (metoclopramide) simply wasn't taken → vomiting. Cisplatin is highly emetogenic with strong DELAYED emesis peaking days 2–3 — the home pills are not optional, and metoclopramide has to actually be taken. ⚠ Poor control in cycle 1 breeds ANTICIPATORY nausea for every future cycle — get it right the first time. Non-drug: small bland meals, hydration, ginger, avoid strong smells.

Filgrastim / pegfilgrastim — stimulates neutrophil recovery. Primary prophylaxis if the regimen's febrile-neutropenia risk is ≥20% (or a high-risk patient); secondary prophylaxis after a febrile-neutropenia episode. Not used to treat uncomplicated neutropenia without fever.
⚠ TIMING — start ≥24 h AFTER the last myelosuppressive dose, never concurrently (G-CSF given during chemotherapy deepens the nadir). With multi-day regimens count from the LAST day (BEP: EP runs d1–5, so G-CSF starts d6 at the earliest).
WHICH ONE — the schedule decides: PEGFILGRASTIM = one 6 mg SC dose per cycle, given ~24–72 h after chemo finishes (not same-day), and it needs ~12–14 days before the next cytotoxic dose → fits q2–3-weekly regimens (incl. dose-dense q14d, given d2). ⚠ NOT suitable for WEEKLY chemotherapy — there is no clear window between doses; use FILGRASTIM instead: daily SC from 24–72 h post-chemo, continued until ANC recovers past the nadir (typically ~5–7 days), stopping ≥24 h before the next dose.
💰 WHEN THE INSURER REFUSES PEGFILGRASTIM — the escalation ladder, in order: ① ask for the BIOSIMILAR — a refusal is usually a formulary brand preference, not a refusal of the class, and the biosimilar is frequently covered when the originator isn't. ② If still refused, substitute DAILY FILGRASTIM ×5–7 days per cycle — same class, same job, far cheaper; the trade is 5–7 injections instead of one. ③ ⚠ WHAT MUST NOT HAPPEN IS A DOSE REDUCTION. If the regimen met the ≥20% threshold, the answer to "no G-CSF" is a different G-CSF, never less chemotherapy — above all in NEOADJUVANT / CURATIVE intent, where dose intensity drives pCR (and pCR vs residual disease then decides adjuvant escalation, e.g. T-DM1 (KATHERINE)). An access problem must not be allowed to become an oncological decision.

Limits: useful for monitoring trends/response, not for diagnosis; confirm progression with imaging, not a single rising value.

What it is: fragments of tumour DNA shed into the bloodstream, detected from a blood test — no tissue needed. Distinct from the protein tumour markers above (CEA/CA125): ctDNA reads the tumour's DNA, so it is far more specific and can be tracked at much lower disease burden.

⚠ Interpretation: POSITIVE = high risk of relapse → investigate/act. NEGATIVE ≠ zero risk — sensitivity varies by assay, and low-shedding tumours + CNS disease are frequently missed (the blood-brain barrier limits shedding). A negative plasma genotype does NOT exclude a mutation → reflex to TISSUE if plasma is negative but suspicion is high. ctDNA complements, does not replace, imaging; interpretation is assay- and tumour-specific.

The drug entries in Part C point here rather than repeat the numbers. General action rule: most toxicities — hold until it recovers to ≤ grade 1, then resume; a grade 3–4 event (or a recurrent grade 2) means dose-reduce on restart, except in curative-intent regimens where you delay-to-recovery + support rather than dose-reduce (germ-cell, many lymphomas). The exact reduction is regimen-specific.

The prognostic anchor comes first — it is what makes the conversation honest. In solid oncology, malignant hepatic dysfunction (rising bilirubin from extensive liver metastases) is one of the strongest short-term prognostic markers there is — median survival typically WEEKS. Add renal impairment and this is multi-organ failure from progressive disease. The conversation is then not "would she want CPR one day" but "she is dying now, and this is how we look after her." Framing it as a hypothetical is the commonest way these conversations fail.

"Dose-dense" = shorter interval between cycles, NOT higher doses. Same drugs, same doses — given every 2 weeks (q14d) instead of every 3 (q21d). Rationale (Norton–Simon): tumour regrows between cycles → giving the next cycle sooner allows less regrowth → more net kill; improved survival in node-positive breast cancer (CALGB 9741). Requires G-CSF every cycle — the marrow can't recover neutrophils in 2 weeks unaided, so growth-factor support is what makes the q2wk interval possible (no G-CSF = no dose-dense). Classic example: dose-dense AC → paclitaxel. If counts still won't recover in the window → dose-reduce or stretch the interval (repeated delays erode the density benefit).

The rule: when the intent is CURE, the dose intensity is part of the cure. In curative regimens — BEP for germ-cell tumours, R-CHOP for lymphoma, curative-intent breast/CRC adjuvant chemo — cutting the dose to get around myelosuppression trades a toxicity problem for a relapse problem. So on the day, if counts are low: ① short DELAY until the ANC (absolute neutrophil count) recovers, then give the full dose; ② add G-CSF secondary prophylaxis (see Drug classes) for the remaining cycles so the next one lands on time; ③ do NOT dose-reduce and do not drop a drug. Contrast with palliative-intent treatment, where the goal is symptom control and quality of life → dose reduction is entirely appropriate to keep the patient on a tolerable therapy. Practical: a low total WBC is not the decision — get the ANC; and afebrile neutropenia is not treated with antibiotics (see Febrile neutropenia above).

A fungating wound is tumour that has broken through the skin — commonest on breast, head and neck, scalp and groin, and usually a sign of neglected, locally advanced or recurrent disease. It is rarely curable by dressings; the definitive treatment of the wound is treatment of the tumour (systemic therapy, or palliative RT, which is the single most useful local measure). Meanwhile four things need managing, and they are not equally urgent. 🔴 BLEEDING is the one that kills — these wounds are friable and can haemorrhage catastrophically. Have a plan before it happens: non-adherent dressings (never rip a dressing off a fungating tumour — soak it off), gentle irrigation rather than swabbing, topical tranexamic acid or adrenaline-soaked gauze / alginate for oozing, haemostatic (single-fraction) radiotherapy or embolisation for repeated bleeds, and dark towels plus a stated plan if a catastrophic bleed is a real possibility. Review anticoagulants and antiplatelets. ODOUR is what isolates the patient socially and it is anaerobic bacteria, not "dirt": topical metronidazole gel (or systemic metronidazole), charcoal-containing dressings, debridement of slough; odour responds, so treat it rather than masking it. EXUDATE and skin damage → absorbent, non-adherent dressings changed on a schedule, barrier film to protect the surrounding skin; keep the dressing change itself covered by analgesia given in advance. INFECTION — these wounds are always colonised, so treat symptoms and signs, not swab results; systemic antibiotics for cellulitis or systemic upset, not for routine colonisation. ⚠ The fifth problem is the one nobody documents: body image and disgust. Patients hide the wound, stop leaving the house, and stop reporting bleeding. Name it, involve the tissue-viability or palliative-care nurses early, and ask specifically about smell and leakage — they will not volunteer it.

The axis: hypothalamus (GnRH) → pituitary (LH/FSH) → ovary → estrogen. Each class blocks a different point — ovarian suppression at the top (premenopausal), aromatase inhibitors at peripheral androgen→estrogen conversion (postmenopausal), and the SERM/SERD at the receptor itself.