Small cell lung cancer (SCLC)

Overview

• Epidemiology: ~13% of all lung ca (was ~25% in 1970s, ↓ d/t ↓ smoking).
• Sex shift: M:F ~50:50 (was 75:25); reflects ↑↑ smoking in women.
• Profile: almost exclusively heavy smokers; central tumors with bulky lymphadenopathy; rapid growth and rapid clinical course.
• Histology: small cells, scant cytoplasm, fine granular chromatin, absent nucleoli, nuclear molding.
• IHC: chromogranin+, synaptophysin+, CD56+, INSM1+, TTF-1+ (often), Ki-67 very high (>70%).
• Molecular subtypes (emerging): SCLC-A (ASCL1), SCLC-N (NEUROD1), SCLC-P (POU2F3), SCLC-I (inflamed). Inflamed subtype may benefit most from IO.
• DLL3 expression: in ~85 to 95% of SCLC, target for tarlatamab and other DLL3-directed therapies.
• Paraneoplastic syndromes (high-yield, see also NSCLC section): SIADH, ectopic ACTH, LEMS (anti-VGCC), cerebellar degeneration (anti-Hu), opsoclonus-myoclonus (anti-Ri), limbic encephalitis (anti-Hu), retinopathy (anti-recoverin), stiff-person (anti-amphiphysin).

Staging

• Limited-stage (LS-SCLC): tumor confined to one hemithorax, with regional nodes that can be encompassed within a single tolerable RT port. ~30% at dx.
• Extensive-stage (ES-SCLC): beyond LS, distant mets, or malignant pleural/pericardial effusion. ~70% at dx.
• TNM staging (per AJCC) is also applied; LS roughly = stages I to III, ES = stage IV.
• Workup:
  • Imaging: CT chest/abdomen/pelvis with contrast; contrast-enhanced brain MRI (high rate of CNS mets); PET-CT (especially LS-SCLC to confirm staging).
  • Labs: CBC, CMP, LDH (prognostic).
  • Bone marrow biopsy: only if unexplained cytopenias or nucleated RBCs suggest marrow involvement (routine marrow exam unnecessary).

Treatment, general principles

• Chemo + RT highly responsive: ORR 60 to 80%, but rapid relapse and chemo-resistance common.
• Poor PS due to SCLC (potentially reversible): still treat with systemic tx, high response rates allow rapid symptom relief. Poor PS not due to SCLC (comorbidity/frailty): individualize; supportive care alone may be appropriate.
• Backbone: platinum (cisplatin or carboplatin) + etoposide × 4 to 6 cycles.
• Cisplatin vs carboplatin: no difference in RR, PFS, or OS; choose per toxicity profile and renal function.
• Cisplatin historically preferred in LS-SCLC (concurrent with RT); carboplatin acceptable substitute.

Limited-stage SCLC (LS-SCLC)

• Standard: cisplatin/etoposide × 4 cycles + concurrent thoracic RT. Start RT with cycle 1 or 2 (early > late). Cisplatin/etoposide is SOC (LS ORR ~97%, CR ~40%), yet 5-yr OS is only ~29 to 34% in modern chemoRT trials (even before consolidation durvalumab).
• Concurrent > sequential: mOS 27.2 vs 19.7 mo (Takada; P=.097, not significant), ↑ esophagitis, ↑ hematologic tox.
• RT regimens (CONVERT trial):
  • 45 Gy BID × 30 fr (3 wk), Turrisi/standard. In Turrisi (NEJM 1999), hyperfractionated (twice-daily) improved mOS 23 vs 19 mo and 5-yr OS 26% vs 16% vs once-daily 45 Gy over 5 wk, at cost of more grade 3 esophagitis. mOS ~30 mo in CONVERT.
  • 66 Gy QD × 33 fr (6.5 wk), not significantly different from BID in CONVERT (superiority design; mOS 25 vs 30 mo, HR 1.18, NS); acceptable alternative if BID logistics not feasible.
  • Higher-dose BID (60 Gy/40 fr), phase 2 data favorable; not yet new SOC.
• Maintenance: prolonged/maintenance chemo has not improved outcomes; radiographic observation is the standard after definitive chemoRT (aside from consolidation durvalumab below).

• ADRIATIC trial (NEJM 2024): durvalumab 1500 mg q4wk × 24 mo as consolidation after concurrent chemoRT in LS-SCLC.
• Efficacy: mOS 55.9 vs 33.4 mo (HR 0.73); mPFS 16.6 vs 9.2 mo (HR 0.76).
• Toxicity: grade 3 to 4 AE 24.4% vs 24.2% (similar to placebo).
• FDA approval: Dec 4, 2024.
• Note: the durvalumab + tremelimumab arm remained blinded/unreported in the 2024 analysis; no added-benefit conclusion can yet be drawn.

• Limited role: consider for clinical T1-2N0 after thorough staging (incl. invasive mediastinal staging). Stage I/II SCLC is rare.
• Post-op: adjuvant chemo recommended regardless of stage (not high-level evidence); if pN+, give adjuvant platinum-etoposide and add mediastinal RT (recommended for N+, with strongest rationale for pN2 or incompletely resected disease; PORT for pure pN1 is less established).

Extensive-stage SCLC (ES-SCLC), first-line

• Baseline chemo activity: cisplatin/etoposide alone gives ORR ~61%, CR ~10%, with fast median time to best response (~4.6 wk), but transient (time to progression ~4 mo, OS ~8.6 mo), underscoring the need for IO and maintenance.
• IMpower133: carboplatin/etoposide +/- atezolizumab × 4 → atezolizumab maintenance.
• Efficacy: mOS 12.3 vs 10.3 mo (HR 0.70); PFS HR 0.77; benefit regardless of PD-L1.
• 5-yr OS: 12% (vs ~2% historical with chemo alone), IMbrella A extension.
• CASPIAN: platinum/etoposide + durvalumab × 4 → durvalumab maintenance, vs platinum/etoposide alone up to × 6 (PCI optional).
• Efficacy: mOS 13.0 vs 10.3 mo (HR 0.73). 3-yr OS 17.6% vs 5.8%.
• KEYNOTE-604: pembrolizumab + chemo, improved PFS but did NOT meet OS endpoint. Negative trial; not used for SCLC.
• Pearl: atezolizumab and durvalumab are interchangeable in ES-SCLC; pick by access/familiarity.

• IMforte trial (Lancet 2025): after 4 cycles of carbo/etoposide + atezolizumab induction, randomize to atezolizumab vs lurbinectedin + atezolizumab maintenance.
• Efficacy: mOS 13.2 vs 10.6 mo (HR 0.73); mPFS 5.4 vs 2.1 mo (HR 0.54).
• FDA approval: Oct 2, 2025 for lurbinectedin (Zepzelca) + atezolizumab (Tecentriq or Tecentriq Hybreza) as 1L maintenance after carbo/etoposide + atezolizumab induction in ES-SCLC. First positive maintenance trial in ES-SCLC.
• Toxicity: grade 3/4 TRAEs 25.6% (vs 5.8% with atezo alone).

Extensive-stage SCLC (ES-SCLC), second-line and beyond

• Definition:
  • Platinum-sensitive: relapse > 6 mo (some say >3 mo) after completing platinum chemo. Rechallenge with platinum/etoposide reasonable.
  • Platinum-resistant: relapse ≤ 3 to 6 mo. Switch agents.

• DeLLphi-301 (phase 2, NEJM 2023): tarlatamab 10 mg q2wk in pretreated SCLC.
• Efficacy: ORR 40%, mDOR 9.7 mo, mOS 14.3 mo.
• DeLLphi-304 (phase 3, NEJM/ASCO 2025): tarlatamab vs investigator-choice chemo (lurbinectedin, topotecan, or amrubicin) post-platinum.
• Efficacy: mOS 13.6 vs 8.3 mo (HR 0.60); 12-mo OS 53% vs 37%.
• Toxicity: grade ≥3 TRAE 27% (vs 62% with chemo). CRS in 56% (grade 3 in 1%), mostly grade 1 to 2, occurs with first 1 to 2 doses.
• Practical: monitor 22 to 24 hrs in an appropriate healthcare setting after C1D1 and C1D8 (not inpatient admission per se); remain within 1 hr of care for 48 hrs with a caregiver (CRS / ICANS risk). Step-up dosing: 1 mg C1 day 1 → 10 mg C1 day 8 and day 15 → then 10 mg q2wk from C2 day 1.
• FDA approval: accelerated May 16, 2024 (DeLLphi-301); full approval Nov 19, 2025 (DeLLphi-304); NCCN Category 1 for 2L+ ES-SCLC.

• Mechanism: alkylating agent (similar to trabectedin used in sarcomas); 3.2 mg/m2 q3wk.
• Phase 2 (105 pts): ORR 35.2%, mOS 9.3 mo overall. CTFI ≥90 days: ORR 45%, mOS 11.9 mo; CTFI <90 days: ORR 22%, mOS 5.0 mo. Excluded brain mets.
• Toxicity: severe myelosuppression (especially neutropenia), nausea, fatigue.
• FDA approval: accelerated Jun 2020. LAGOON phase 3 confirmatory trial ongoing.

• Topotecan: prolongs OS vs best supportive care (older trial). Often poorly tolerated (cytopenias). Now largely superseded by tarlatamab/lurbinectedin.
• Re-platinum (cisplatin/etoposide or carbo/etoposide): for platinum-sensitive relapse > 6 mo.
• Single-agent chemo: irinotecan, paclitaxel, gemcitabine, vinorelbine, temozolomide (TMZ not FDA-approved for SCLC but NCCN-listed).
• Investigational: EZH2 and PARP1 inhibitors under study.
• Withdrawn IO indications:
  • Nivolumab (3L monotherapy ORR ~11.9%): SCLC indication withdrawn 2020.
  • Pembrolizumab (3L monotherapy ORR ~19%): SCLC indication withdrawn 2021.
  • Only approved IO use in SCLC is 1L (with platinum/etoposide) and post-chemoRT consolidation (durvalumab); however, both nivolumab and pembrolizumab remain in NCCN to allow relapsed pts to receive IO if not given in 1L.

Prophylactic cranial irradiation (PCI)

• LS-SCLC with CR/PR to chemoRT:
• PCI recommended, Auperin meta-analysis: ↓ brain mets, 5.4% absolute ↑ in 3-yr OS.
• ES-SCLC with response to induction:
• Older EORTC trial (Slotman 2007): PCI ↓ brain mets, ↑ mOS 6.7 vs 5.4 mo, 1-yr OS 27.1% vs 13.3%, 1-yr brain mets 14.6% vs 40.4%. Limitation: no baseline brain MRI required.
• Modern Japanese trial (Takahashi 2017, w/ MRI surveillance, 224 pts): no OS benefit (mOS 11.6 with PCI vs 13.7 mo observation). Many now favor MRI surveillance over PCI. SWOG/MAVERICK trial ongoing to clarify.
• In durvalumab era: PCI use is decreasing, patients on ADRIATIC could receive PCI per investigator (used in ~54%); MRI surveillance is a reasonable alternative.
• PCI dose: 25 Gy in 10 fractions standard. Higher doses (e.g., 36 Gy) → ↑ neurocognitive tox without OS benefit.
• Toxicity: cognitive decline, fatigue. Hippocampal-sparing PCI may reduce neurocog decline (NRG CC003).

Consolidative thoracic RT in ES-SCLC

• Slotman CREST trial (pre-IO era): in ES-SCLC responders to platinum-based chemo, thoracic RT added to PCI did NOT meet 1-yr OS endpoint (33% vs 28%, P=.066), but secondary 2-yr OS 13% vs 3% (P=.004); reduced intrathoracic recurrence.
• Modern era: role unclear given chemo-IO +/- maintenance is standard. Consider case-by-case (especially residual thoracic disease).

High-yield SCLC pearls

• LS-SCLC standard: platinum/etoposide × 4 + concurrent thoracic RT (45 Gy BID or 66 Gy QD) → consolidation durvalumab × 24 mo (ADRIATIC).
• ES-SCLC 1L: carbo/etoposide + atezolizumab × 4 → atezolizumab +/- lurbinectedin maintenance (IMforte); OR platinum/etoposide + durvalumab × 4 → durvalumab maintenance (lurbinectedin not approved with durvalumab).
• ES-SCLC 2L: tarlatamab (DLL3 BiTE) is new SOC after platinum (DeLLphi-304).
• CRS with tarlatamab: 51% (DeLLphi-301) to 56% (DeLLphi-304), mostly G1 to 2; monitor 22 to 24 hrs in an appropriate healthcare setting after the first 2 doses (C1D1, C1D8).
• PCI: still recommended in LS-SCLC after CR; ES-SCLC role debated in MRI-surveillance era.
• 45 Gy BID over 3 weeks remains the LS-SCLC RT standard (CONVERT did not establish QD superiority).
• Nivolumab/pembrolizumab SCLC monotherapy indications were WITHDRAWN, IO is for 1L (induction + maintenance) and post-CRT (durvalumab).
• DLL3: expressed on ~85 to 95% of SCLC; target for tarlatamab.
• Subtypes (SCLC-A/N/P/I), emerging biomarkers; SCLC-I may be most IO-responsive.
• Brain MRI is mandatory at staging and during follow-up (high CNS met rate; replaces PCI in some practices).
• Paraneoplastic LEMS = anti-VGCC antibodies (proximal weakness improving with repeated effort), contrast with myasthenia gravis.