Lung Cancer: Matching the Drug to the Tumour's DNA
Two patients, two chest scans that look almost identical — yet one goes home with a single pill and the other starts an intravenous drip of immunotherapy plus chemotherapy. What separated them was not the X-ray but the molecular report on the biopsy. Lung cancer is where oncology stopped treating an organ and started treating a genotype. This chapter assembles everything from the earlier drug-class chapters and shows how the pieces fit one disease.
Two beds, side by side. On both scanners a spiculated mass sits in the right upper lobe — to the eye the tumours are twins. A 61-year-old lifelong non-smoker in bed one; a 68-year-old with forty pack-years in bed two. Same size, same lobe, same word on the report: adenocarcinoma. Yet a week later she is swallowing one small tablet each morning at home, and he is hooked to an infusion of an antibody plus platinum chemotherapy in the day unit. Nothing on the X-ray explained the gap. The biopsy's molecular panel did: her tumour carried an EGFR mutation, his carried none. The scan showed the shadow; the DNA wrote the prescription.
The first fork: two diseases wearing one name
"Lung cancer" is not one disease — it is two, and they diverge almost immediately. Under the microscope the pathologist first sorts every lung tumour into non-small-cell lung cancer (NSCLC), which is about 85% of cases, or small-cell lung cancer (SCLC), about 15%. This is not a bookkeeping detail; it is the single most treatment-changing split in the whole field. NSCLC (adenocarcinoma, squamous cell, large cell) is the home of precision oncology and immunotherapy. SCLC is aggressive, doubles fast, is almost always tied to heavy smoking, spreads early — and is fought mainly with chemotherapy. Get this fork wrong and every downstream decision is wrong.
A memory hook: NSCLC is the "read-the-DNA-first" cancer; SCLC is the "start-chemo-fast" cancer. NSCLC gives you time and choices because a targetable driver or a high PD-L1 may hand you a far better first option than chemotherapy. SCLC rarely does — it is extraordinarily chemo-sensitive up front but relapses, so the strategy is speed, not selection.
NSCLC: test before you treat
For advanced non-squamous NSCLC the modern rule is blunt: do not write a single chemotherapy order until two questions are answered. First — does the tumour carry a targetable driver mutation? Second — how much PD-L1 does it express? These come from the biopsy, run as a molecular panel and an immunohistochemistry stain (the same biomarker-testing logic laid out in the Precision-oncology chapter). Only when the report returns do you know which of three very different roads this patient is on: a targeted pill, immunotherapy, or chemotherapy — usually not chosen by the doctor's preference but dictated by the tumour's biology.
An activating EGFR mutation (common in non-smokers, women, adenocarcinoma) is answered with an EGFR tyrosine-kinase inhibitor. Osimertinib is the preferred first-line agent; erlotinib and gefitinib are the older members of the class. These are oral drugs the patient takes at home — the mechanism, an ATP-competitive block of the receptor's kinase domain, is detailed in the Kinase-inhibitors chapter. This is the woman in bed one: her whole treatment is one tablet a day.
An ALK gene rearrangement (again typical of never-smokers) is matched to an ALK inhibitor — alectinib is preferred first-line, with crizotinib the original agent (crizotinib also covers ROS1 rearrangements). Same principle as EGFR, different lock and key: find the fusion, hand over the pill built to fit it. Both EGFR and ALK inhibitors belong to the kinase-inhibitor family covered earlier.
The driver list keeps growing, and so does the toolbox. Beyond EGFR and ALK, testing now routinely looks for several more actionable alterations, each with a matched inhibitor: BRAF V600E (a BRAF plus MEK inhibitor combination, the same drugs met in melanoma), KRAS-G12C (sotorasib, the first drug to crack a target once called "undruggable"), MET exon-14 skipping, and RET fusions (selective RET inhibitors). The practical lesson is not the drug names but the reflex: in NSCLC you send a broad panel, because any one positive result can replace toxic chemotherapy with a targeted pill.
- Split every lung tumour first: NSCLC (~85%) vs SCLC (~15%). This decides everything.
- In advanced NSCLC, test driver mutations + PD-L1 BEFORE treating.
- EGFR mutation → osimertinib (or erlotinib/gefitinib), an oral kinase inhibitor.
- ALK/ROS1 rearrangement → alectinib/crizotinib.
- BRAF, KRAS-G12C (sotorasib), MET, RET each have a matched inhibitor — send a broad panel.
No driver? Then PD-L1 and immunotherapy lead
Many tumours — including most in smokers — carry no targetable driver. Here the second number takes over: PD-L1. A tumour expressing high PD-L1 is one that has been hiding from the immune system by pressing the PD-1 "off switch" on T cells; a checkpoint inhibitor releases that brake. Pembrolizumab, an anti-PD-1 antibody, is the workhorse — as single-agent first-line therapy when PD-L1 is high, and combined with chemotherapy when it is lower or the tumour burden is large. The mechanism — antibodies that block the PD-1/PD-L1 checkpoint so T cells re-engage the tumour — is the subject of the Checkpoint-inhibitors chapter.
When chemotherapy is used — alone or, more often now, alongside immunotherapy — the backbone is a platinum doublet: a platinum agent (cisplatin or carboplatin) paired with a second cytotoxic. In non-squamous NSCLC that partner is usually pemetrexed; in squamous disease a taxane such as paclitaxel. In non-squamous tumours the anti-angiogenic antibody bevacizumab is often added to starve the tumour's blood supply. Each of these is a returning face: platinums and pemetrexed and the taxanes from the cytotoxic chapters, bevacizumab from the Monoclonal-antibodies chapter. Lung cancer does not invent new drugs so much as choreograph the ones you already know.
- No driver mutation → PD-L1 guides therapy; high PD-L1 favours single-agent pembrolizumab.
- Lower PD-L1 or heavy burden → pembrolizumab + platinum-doublet chemotherapy.
- Platinum doublet = cisplatin/carboplatin + pemetrexed (non-squamous) or a taxane (squamous).
- Bevacizumab is often added in non-squamous tumours to block angiogenesis.
- Immunotherapy checkpoint blockade is the biggest single advance in driver-negative NSCLC.
SCLC: fast, chemo-sensitive, and full of surprises
Small-cell lung cancer plays by opposite rules. There is no precision-oncology menu here — driver-targeted pills have not delivered in SCLC, so the reflex biomarker hunt of NSCLC is largely skipped. Instead the disease is treated for what it is: fast-growing, widely metastatic at diagnosis, and (paradoxically) very sensitive to chemotherapy at first. First-line therapy is a platinum agent plus etoposide, now usually combined with a checkpoint inhibitor such as atezolizumab. Responses are often dramatic — tumours shrink quickly — but the catch defines the disease: it almost always relapses, and relapsed SCLC is far harder to treat. The etoposide here is the topoisomerase inhibitor from the cytotoxic chapters; atezolizumab is another anti-PD-L1 checkpoint antibody.
SCLC also has a signature that NSCLC rarely shows: paraneoplastic syndromes. These neuroendocrine tumours secrete hormones and provoke antibodies that cause disease at a distance from the tumour itself. Three are classic: SIADH (inappropriate ADH → hyponatraemia), ectopic ACTH (→ Cushing's syndrome), and Lambert-Eaton myasthenic syndrome (antibodies against calcium channels → proximal muscle weakness that improves with use). A patient whose low sodium or new weakness turns out to be the first clue to a hidden lung tumour is a scenario worth recognising — the endocrine mechanics of SIADH are covered in the Endocrine section.
Smoking drives the overwhelming majority of lung cancers, and it is not merely history-taking trivia — smoking cessation is itself a therapeutic intervention. Continued smoking worsens treatment tolerance, response, and survival even after diagnosis, so nicotine-replacement therapy, varenicline or bupropion are offered as part of care. The pharmacology of cessation belongs to the Respiratory section; the point here is that in lung cancer, stopping the cause is written into the treatment plan, not left as advice.
- SCLC: skip the driver hunt — treat fast with platinum + etoposide ± atezolizumab.
- SCLC is very chemo-sensitive at first but relapses; relapse is hard to treat.
- Think paraneoplastic in SCLC: SIADH, ectopic Cushing's, Lambert-Eaton.
- Smoking cessation is treatment, not just advice — it improves outcomes after diagnosis.
- Biomarker testing now precedes treatment in NSCLC; it is not (yet) actionable in SCLC.
- Starting empirical chemotherapy in advanced NSCLC before biomarker and PD-L1 testing — you may miss a far more effective targeted or immunotherapy option.
- Treating NSCLC and SCLC as one disease. NSCLC is targetable/immunotherapy-led; SCLC is chemo-driven — the strategies are opposite.
- Forgetting paraneoplastic syndromes in SCLC. A hyponatraemia, a Cushingoid picture or unexplained proximal weakness may be the tumour's first signal.
- Assuming a high PD-L1 result means the patient should skip chemotherapy entirely — it favours immunotherapy first-line but the regimen still depends on stage and burden.
A 60-year-old never-smoker has advanced NSCLC adenocarcinoma. The molecular panel reports an activating EGFR mutation. The most appropriate first-line therapy is:
- First split every lung cancer: NSCLC (~85%, targetable/immune) vs SCLC (~15%, chemo-driven).
- In NSCLC, test driver mutations + PD-L1 first: EGFR→osimertinib, ALK/ROS1→alectinib, plus BRAF/KRAS-G12C/MET/RET.
- No driver → immunotherapy (pembrolizumab, esp. high PD-L1) ± platinum doublet ± bevacizumab.
- SCLC → platinum + etoposide ± atezolizumab; chemo-sensitive but relapses; watch paraneoplastic syndromes.
- The biopsy's molecular report — not the scan — writes the modern lung-cancer prescription; smoking cessation is part of treatment.
- Katzung BG. Basic & Clinical Pharmacology — Cancer Chemotherapy: targeted agents, kinase inhibitors & immunotherapy in NSCLC/SCLC.
- NCCN Clinical Practice Guidelines in Oncology — Non-Small Cell Lung Cancer (biomarker testing, EGFR/ALK/ROS1/BRAF/KRAS-G12C/MET/RET, PD-L1).
- NCCN Clinical Practice Guidelines in Oncology — Small Cell Lung Cancer (platinum-etoposide, immunotherapy).
- Planchard D, et al. ESMO Clinical Practice Guidelines — Metastatic non-small-cell lung cancer.
- Brunton LL, et al. Goodman & Gilman's The Pharmacological Basis of Therapeutics — Targeted therapies & immune checkpoint inhibitors.
