Thoracic Malignancies — Board Review Summary

High-yield review of lung cancer screening, AJCC v9 updates, early-stage and locally advanced NSCLC, postoperative RT, and small cell lung cancer, with emphasis on board-relevant dose/fractionation, immunotherapy transitions, and modern toxicity lessons.

PART I — EPIDEMIOLOGY, SCREENING, AJCC v9

Screening — NLST and NELSON

Trial	Population	Intervention	Result
NLST (NEJM 2011)	53,454 pts, age 55-74, ≥30 pack-yr, current or quit <15 yr	Annual LDCT x3 vs CXR	20% relative reduction in lung cancer mortality; 6.7% reduction in all-cause mortality
NELSON (NEJM 2020)	13,195 men, 2,594 women; age 50-74, ≥15 cig/day for ≥25 yr or ≥10/day for ≥30 yr	LDCT at 0, 1, 3, 5.5 yr vs no screening	10-yr lung cancer mortality reduced 24% in men, 33% in women

USPSTF 2021 criteria (current, widely adopted): ages 50-80, ≥20 pack-yr, currently smoking or quit within 15 yr. Using strict NLST criteria would only capture about 27% of incident lung cancers; broader eligibility improves yield.

AJCC v9 — Key Changes (memorize)

Category	Subdivision in v9	Rationale
N2 — ipsilateral mediastinal/subcarinal	N2a = single station; N2b = multiple stations	Single-station N2 has better prognosis in surgical series (Int-0139: HR 0.72)
M1c — distant mets	M1c1 = multiple extrathoracic mets, single organ system; M1c2 = multiple organ systems	Reflects the growing distinction between limited-burden and more disseminated metastatic disease

T and M1a/M1b are otherwise unchanged from v8. M1a: contralateral lung nodule, pleural/pericardial nodules, malignant pleural/pericardial effusion. M1b: single extrathoracic metastasis in a single organ.

PART II — EARLY-STAGE NSCLC

SBRT vs Surgery: STARS / ROSEL Pooled (Chang Lancet Oncol 2015; revised Lancet Oncol 2021)

58 operable pts T1-T2a N0 NSCLC pooled from two closed phase III trials. STARS: CyberKnife 18 Gy x 3 peripheral, 12.5 Gy x 4 central. ROSEL: linac SBRT 18 Gy x 3 or 12 Gy x 5, peripheral only. 3y OS 95% SBRT vs 79% lobectomy (p=0.037). Updated 2021 revised analysis with longer follow-up: SBRT 5 recurrences / 1 death vs lobectomy 6 recurrences / 6 deaths. Suggestive, but not definitive — both trials closed early because of poor accrual.

Operable T1-T2 NSCLC: STABLE-MATES / VALOR ongoing

Phase III comparisons are still pending. For clearly operable early-stage disease, the reference standard remains resection with mediastinal nodal evaluation. JCOG 0802 strengthened the case for segmentectomy in selected peripheral tumors up to 2 cm, but this does not make SBRT and surgery interchangeable.

Biopsy Before SBRT?

ASTRO guideline: biopsy whenever possible to confirm histologic diagnosis. If pretest probability >85% on a validated model (e.g., Mayo, Brock), empiric SBRT is acceptable. Two large cohorts (676 pts, 65% no-bx; 701 pts, 33% no-bx) showed no significant difference in local, regional, or distant outcomes.

Does Dose Matter? The sBED Concept

"Size-adjusted BED" (sBED) = BED − tumor diameter (mm). Achieving sBED >= 80 Gy correlates with about 90-95% local control. Examples: 10 Gy x 5 for a 2 cm tumor or 12 Gy x 5 for a 5 cm tumor. Single-fraction SBRT and multi-fraction high-BED regimens can produce comparable control when the effective dose is adequate.

Landmark SBRT Dose-Response Trials

Trial	Design	Result
RTOG 0236 (Timmerman JAMA 2010)	T1-T2 peripheral, medically inoperable; 18 Gy x 3 (60 Gy)	3y primary tumor control 98%; 3y LC 91%; 3y OS 56%
RTOG 0618 (Timmerman 2018)	T1-T2 peripheral, operable; 18 Gy x 3	4y LC 96%, OS 56%; confirms strong SBRT efficacy in operable pts, though not as a surgery-comparison trial
CHISEL (Ball Lancet Oncol 2019)	RCT: SBRT vs conventionally fractionated RT, T1-T2a peripheral inoperable	SBRT reduced LR (HR 0.32) and improved OS (HR 0.53)

Central / Ultra-central SBRT — High-Yield Spectrum

Trial	Target / Dose	Toxicity
Timmerman "No-Fly Zone" (JCO 2006)	Central tumors: 20-22 Gy x 3	G3-5 toxicity unacceptably high; 3-fx regimens should NOT be used near central airways
RTOG 0813 (Bezjak JCO 2019) — dose finding	Central, within 2 cm of proximal bronchial tree OR PTV touching mediastinal/pericardial pleura. 10-12 Gy x 5	12 Gy x 5 = MTD (7% DLT risk at 1 yr). DLTs included G3 hypoxia, G3 pneumonitis, G5 bradycardia, and one death NOS
HILUS (Lindberg JTO 2021)	Primary or metastatic lung tumor <5 cm within 1 cm of proximal bronchial tree. 7 Gy x 8 to 67% isodose, 4 fx/week	G5 toxicity, mostly hemoptysis: 21% near trachea/mainstem vs 8% near lobar/intermedius bronchi
Follow-up analysis	7 Gy x 8 pts within 2 cm of PBT (including HILUS)	Bronchus intermedius should be treated with the same caution as the mainstem bronchi
SUNSET (Giuliani IJROBP 2024)	Ultra-central primary tumors <6 cm with PTV overlapping PBT, esophagus, pulm vein, or pulm artery. 7.5 Gy x 8, hot spot <120%, daily fx	1/30 (3%) G5 toxicity, a markedly safer result than historical HILUS-style 7 Gy x 8 experience

Ohri's practical SBRT framework (memorize):

Peripheral tumors: 18 Gy x 3. Switch to 12 Gy x 5 if chest wall constraints cannot be met.
Ultra-central (within 1 cm of trachea, mainstem bronchus, bronchus intermedius, or esophagus): 7.5 Gy x 8 (SUNSET).
Other central tumors: 10 Gy x 5 (<2-3 cm) or 12 Gy x 5 (>2-3 cm), while respecting 10/5-style constraints.
Aim for 120-130% hot spots.
Central = within 2 cm of PBT OR PTV touching mediastinal/pericardial pleura.

SBRT Contraindications — ILD

Pretreatment PFTs do NOT reliably predict pneumonitis risk; the largest decline may occur in pts with the best baseline function. Small fibrotic ILD series treated with 50 Gy/5 fx showed 8% G4-5 toxicity (all pulmonary deaths) — caution is essential.

SBRT + Immunotherapy for Early-Stage NSCLC

Phase II I-SABR suggested activity for SBRT + nivolumab. Phase III trials have not confirmed a role: SWOG 1914 (atezo) closed for futility, KEYNOTE 867 (pembro) closed for futility, and PACIFIC-4 has completed accrual. As of 2026, there is no phase III evidence supporting routine IO after SBRT for early-stage NSCLC.

PART III — LOCALLY ADVANCED NSCLC

Foundational Trials

Trial	Design	Result
RTOG 73-01 (Perez 1980s)	Dose-finding RT alone: 40 vs 50 vs 60 Gy	Trend favoring 60 Gy. Median OS about 1 yr in an unresectable, pre-modern-systemic-therapy era
Sequential CRT meta-analysis	11 trials, cisplatin-based, 50-65 Gy, mostly induction chemo	3% absolute OS benefit at 2y (HR 0.90)
Concurrent CRT meta-analysis	6 trials, 49-66 Gy; platinum-based	5% absolute OS benefit at 5y (HR 0.84). Median OS about 18 mo. Established concurrent CRT as SOC
Elderly subgroup (JCOG 0301)	Age >70, cisplatin-ineligible; 60 Gy +/- low-dose carbo	Concurrent low-dose carboplatin improved OS in fit elderly pts
Poor-prognosis (CONCERT/TROG)	Tumor ≥8 cm and/or PS ≥2 and/or weight loss >10%. Carbo+vinorelbine +/- RT 42 Gy/15 fx	RT benefit was more modest in a frailer population

RTOG 0617 (Bradley JCO 2020 long-term) — The Seminal Negative Dose-Escalation Trial

2x2 design: standard 60 Gy vs high-dose 74 Gy; +/- cetuximab. Carbo/taxol concurrent. The 74 Gy arm had worse OS (median 20 vs 28.7 mo). Cetuximab provided no benefit. Mean heart dose emerged as a major adverse predictor: HR 1.03/Gy for symptomatic cardiac events and HR 1.07/Gy for G3+ cardiac events. This trial cemented 60 Gy as the standard dose and made cardiac sparing central to plan quality.

Failed Attempts to Improve Concurrent CRT (all negative RCTs)

Consolidation chemotherapy (HOG, KCSG)
Maintenance targeted therapy (SWOG S0023 — gefitinib, inferior OS)
Prophylactic cranial RT (RTOG 0214)
Trimodality (INT 0139 — no OS benefit overall)
Dose escalation (RTOG 0617 — inferior OS)
Concurrent cetuximab (RTOG 0617)
New chemotherapy platforms (PROCLAIM — pem/cis)
Metformin (NRG-LU001)
PET-based dose escalation (RTOG 1106, PET-PLAN)

Modern Constraints for LA-NSCLC CRT (Ohri)

Lung: V20 <35% historically; composite EQD2 V20 <40%. Mean lung dose as low as feasible.
Esophagus: minimize composite EQD2 max dose above 100 Gy; avoid above 110 Gy; strongly discourage above 120 Gy.
Heart: mean heart dose ALARA; heart V40 <50%.

PACIFIC (Antonia NEJM 2017; Spigel JCO 2022 5-yr)

713 pts with unresectable stage III NSCLC without progression after platinum-based CRT, randomized within 42 days of CRT completion. Durvalumab x 12 months vs placebo. Median PFS 16.9 vs 5.6 mo; median OS 47.5 vs 29.1 mo. 5y OS 42.9% vs 33.4%. PACIFIC reduced intrathoracic progression, distant failure, and brain metastases, and it remains the defining consolidation trial for CRT-responsive wild-type stage III disease.

PACIFIC caveats: do NOT use durvalumab for EGFR-mutant or ALK-rearranged tumors when a targeted post-CRT strategy exists; benefit in PD-L1 TPS 0% remains less certain; and roughly 20-25% of real-world CRT pts never reach consolidation because of toxicity or early progression.

Concurrent vs Adjuvant IO + CRT

Trial	Design	Result
PACIFIC-2 (Bradley 2024)	Concurrent + adjuvant durvalumab vs placebo, with CRT	Negative for PFS; no OS benefit
EA5181 (ongoing)	Concurrent + consolidative durvalumab vs consolidative alone	Primary endpoint OS; pending

LAURA (Lu NEJM 2024) — EGFR-Mutant Unresectable Stage III

216 pts with EGFR-mutant (ex19del or L858R) unresectable stage III NSCLC without progression after definitive CRT. Osimertinib 80 mg daily until progression vs placebo. Median PFS 39.1 vs 5.6 mo (HR 0.16, p<0.001). Osimertinib is now SOC for EGFR+ unresectable stage III after CRT and should replace durvalumab in this subgroup.

RT + IO Without Chemo (for CRT-Ineligible)

Trial	IO / PD-L1 requirement	RT regimen
NRG-LU004	Durvalumab x 1y (started with RT); PD-L1 ≥50%	60 Gy/30 fx or 60 Gy/15 fx
SPRINT (Montefiore)	Pembrolizumab q3wk x3 before RT, then x13 after; PD-L1 ≥50%	48/55 Gy/20 fx PET-based dose-painted
DOLPHIN (Japan)	Durvalumab x 1y (with RT); PD-L1 ≥1%	60 Gy/30 fx
SWOG S1933	Atezolizumab x 1y (after RT); any PD-L1	60 Gy/15 fx
DART (MSKCC)	Durvalumab x 1y (with RT); any PD-L1	60 Gy/30 fx
TRADE-hypo (Germany)	Durvalumab x 1y (with RT); any PD-L1	60 Gy/30 fx or 55 Gy/20 fx (closed for toxicity)
AIRING (France)	Nivolumab x 6 mo (with RT); any PD-L1	66 Gy/24 fx

Surgery for LA-NSCLC — Key Negative Trials

Trial	Design	Result
INT 0139 (Albain Lancet 2009)	396 pts stage IIIA(N2) resectable: CRT 45 Gy + cis/etop → resection vs definitive CRT 61 Gy	PFS benefit with surgery but no OS difference. Exploratory signal favored lobectomy, not pneumonectomy. Single-station N2 was prognostically favorable but not clearly predictive of surgery benefit
EORTC 08941 (van Meerbeeck JNCI 2007)	333 stage IIIA(N2) unresectable after 3 cycles chemo responders: resection vs RT about 60 Gy	No OS or PFS difference; surgery carried higher treatment-related mortality

Perioperative Immunotherapy (CheckMate 816, NADIM II, KEYNOTE-671, AEGEAN)

CheckMate 816 (Forde NEJM 2022): neoadjuvant nivo + platinum-doublet chemo vs chemo alone (resectable IB-IIIA). pCR 24% vs 2.2%; EFS HR 0.63; MPR 36.9% vs 8.9%.
NADIM II: neoadjuvant nivo + chemo for resectable IIIA disease: pCR 36.8% vs 6.9%.
KEYNOTE-671 and AEGEAN: peri-operative pembro and durva, respectively, improved EFS in resectable stage II-IIIB disease.
Practical takeaway: if a patient receives induction chemo-IO but does not proceed to surgery, management usually reverts to a definitive CRT framework followed by the appropriate post-CRT systemic strategy.

PART IV — POSTOPERATIVE (PORT) NSCLC

PORT Meta-Analysis (Lancet 1998; updated 2005)

Adjuvant RT after complete resection for NSCLC produced a 7% absolute decrement in 2y OS in pN0/pN1. The apparent harm was driven largely by outdated techniques (2D planning, cobalt-era treatment). Benefit in pN2 remained uncertain.

Lung ART (Le Pechoux Lancet Oncol 2022) — Modern PORT for pN2

501 pts with completely resected pN2 NSCLC randomized to PORT 54 Gy/27-30 fx vs observation. 3y DFS 47% vs 44% (HR 0.86, NS). OS HR 0.97 with more cardiopulmonary toxicity in the PORT arm. Modern PORT did not improve DFS or OS.

Current takeaway: PORT for pN2 is no longer routine. Consider it selectively for unusually high-risk scenarios such as positive/close margins, extracapsular extension, incomplete nodal dissection, or concerning residual mediastinal findings.

PART V — SMALL CELL LUNG CANCER

Limited-Stage SCLC (LS-SCLC)

Classic BID RT: Intergroup 0096 (Turrisi NEJM 1999)

Concurrent cis/etoposide + RT: 45 Gy BID (1.5 Gy x 30 fx over 3 wk) vs 45 Gy daily (1.8 Gy x 25 fx over 5 wk). 5y OS 26% vs 16% favoring BID. Local failure 36% vs 52%. G3 esophagitis was higher with BID but the survival gain made BID the classic benchmark regimen.

Dose-Escalated Daily RT (Equivalent Options)

Trial	Arms	Result
CONVERT (Faivre-Finn Lancet Oncol 2017)	66 Gy/33 fx daily vs 45 Gy BID	Equivalent OS (2y OS 56% vs 51%); similar G3+ esophagitis
CALGB 30610 / RTOG 0538 (Bogart 2023)	70 Gy/35 fx daily vs 45 Gy BID	Equivalent OS; trend toward improved local control with 70 Gy, without a definitive survival advantage

"Very Limited" Stage SCLC (T1-2 N0)

Lobectomy with mediastinal staging is reasonable in carefully selected peripheral node-negative disease. Adjuvant chemotherapy is standard; thoracic RT is added if nodal disease is found. SBRT-only strategies remain investigational.

ADRIATIC (Cheng NEJM 2024) — Practice Change for LS-SCLC

730 pts with LS-SCLC without progression after concurrent CRT randomized to durvalumab (+/- tremelimumab) x 24 mo vs placebo. Median OS 55.9 vs 33.4 mo (HR 0.73); median PFS 16.6 vs 9.2 mo. Adjuvant durvalumab after CRT is now SOC.

NRG-LU005 (Higgins 2024) — Concurrent Atezolizumab with CRT

544 LS-SCLC pts randomized to CRT + atezolizumab (concurrent + consolidation) vs CRT alone. Negative for both OS and PFS, echoing the broader pattern that adjuvant sequencing has outperformed concurrent IO in thoracic CRT settings.

Extensive-Stage SCLC (ES-SCLC)

First-line systemic (IMpower133, CASPIAN): platinum/etoposide + atezolizumab or durvalumab. Median OS about 12-13 mo.
Consolidative thoracic RT (Slotman Lancet 2015): in chemo-responders, thoracic RT 30 Gy/10 fx improved longer-term thoracic control and 2y OS on secondary analysis.
In the IO era, consolidative thoracic RT remains a selective tool for pts with persistent intrathoracic bulky or symptomatic disease.

Prophylactic Cranial Irradiation — The Evolving Story

Trial	Population	Finding
Auperin (NEJM 1999, meta-analysis)	LS-SCLC CR after CRT	About 5% absolute OS benefit at 3y; major reduction in brain metastases
Le Pechoux (Lancet Oncol 2009) — PCI dose	720 LS-SCLC CR pts: 25 Gy/10 fx vs 36 Gy	25 Gy remained standard; 36 Gy was not better and was associated with worse OS
Slotman (EORTC 22993-08993, NEJM 2007)	286 ES-SCLC responders: PCI vs observation. No baseline MRI required	1y symptomatic brain mets 15% vs 40%; 1y OS 27% vs 13%
Takahashi (Lancet Oncol 2017)	224 ES-SCLC responders: PCI vs MRI surveillance with mandatory baseline and serial MRI	Negative for OS; changed modern practice toward MRI surveillance rather than routine PCI in ES-SCLC

Hippocampal Avoidance PCI (HA-PCI)

Trial	Design	Finding
GOECP-SEOR (Spain)	150 pts; PCI 25 Gy +/- HA. Primary: FCSRT delayed free recall decline	Positive: 6% vs 24% decline (p=0.003), without loss of intracranial control
NKI (Netherlands)	168 pts; PCI 25 Gy +/- HA. Primary: HVLT total recall decline	Negative; likely affected by contouring/QC limitations
NRG-CC003 (Gondi 2024)	393 pts, 53% memantine use; HA-PCI vs PCI	Primary HVLT-R delayed recall endpoint NS, but composite neurocognitive failure favored HA-PCI (aHR 0.78)

HA-PCI summary: mixed phase III data, but enough signal to make HA-PCI reasonable when contouring is rigorous and memantine is used. Per NRG-CC003, hippocampal D100% should remain <9 Gy and Dmax <16 Gy.

PART VI — CROSS-CUTTING HIGH-YIELD POINTS

AJCC v9: N2 now splits into N2a (single station) and N2b (multiple stations); M1c splits by single-organ-system vs multi-organ-system dissemination.
Screening: NLST showed a 20% lung cancer mortality reduction; NELSON confirmed major benefit in a European cohort. USPSTF 2021 uses ages 50-80, ≥20 pack-yr, quit <15 yr.
SBRT dose/fractionation: peripheral 18 Gy x 3; central 12 Gy x 5; ultra-central 7.5 Gy x 8. Avoid 3-fx regimens near central airways.
HILUS matters: 7 Gy x 8 near trachea/mainstem carried very high fatal hemoptysis risk.
sBED >= 80 Gy correlates with excellent local control in peripheral SBRT.
Biopsy before SBRT is preferred unless the malignancy probability is convincingly high and tissue cannot be safely obtained.
RTOG 0617 established that 74 Gy is worse than 60 Gy and highlighted the importance of heart dose.
PACIFIC remains the standard post-CRT strategy for unresectable wild-type stage III NSCLC.
LAURA is the new standard for EGFR+ unresectable stage III disease after CRT.
Concurrent IO + CRT has not outperformed the adjuvant/consolidation approach in either NSCLC or LS-SCLC.
Lung ART means PORT is no longer routine after complete resection of pN2 NSCLC.
Turrisi BID 45 Gy remains a classic LS-SCLC standard; CONVERT and CALGB 30610 support daily alternatives.
ADRIATIC made adjuvant durvalumab standard after LS-SCLC CRT.
ES-SCLC PCI is now optional in the MRI-surveillance era because Takahashi failed to confirm the earlier EORTC survival benefit.
HA-PCI is reasonable with strong QA and memantine, but the randomized data are mixed.
Trimodality in stage IIIA(N2) is not a routine OS-improving strategy; lobectomy-only subgroups remain hypothesis-generating.
Perioperative IO is now established in resectable NSCLC, but if surgery does not happen, management generally falls back to definitive CRT principles.

KEY LANDMARK TRIALS (memorize)

Trial	Disease	One-line takeaway
NLST (2011)	Lung screening	LDCT reduced lung cancer mortality by 20%
NELSON (2020)	Lung screening (Europe)	Confirmed a major mortality benefit, especially in women
STARS/ROSEL pooled (2015/2021)	Operable T1-T2 N0 NSCLC	Suggestive SBRT survival signal, but underpowered and not definitive
RTOG 0236 (2010)	Inoperable peripheral T1-T2 NSCLC	Established high local control with 18 Gy x 3
RTOG 0618 (2018)	Operable peripheral T1-T2 NSCLC	Confirmed SBRT efficacy in operable pts
CHISEL (2019)	Peripheral inoperable T1-T2a	SBRT outperformed conventional RT
RTOG 0813 (2019)	Central T1-T2 NSCLC	12 Gy x 5 identified as the MTD
HILUS (2021)	Ultra-central lung tumors	Highlighted major fatal bleeding risk with unsafe 8-fx schedules
SUNSET (2024)	Ultra-central NSCLC	7.5 Gy x 8 provided a safer modern approach
RTOG 73-01	LA-NSCLC RT alone	Helped establish 60 Gy as the classic RT-alone benchmark
Concurrent CRT meta-analysis	LA-NSCLC	Concurrent CRT improved OS over sequential therapy
RTOG 0617	LA-NSCLC	74 Gy was worse than 60 Gy; heart dose matters
PACIFIC	Unresectable stage III NSCLC	Durvalumab after CRT improved PFS and OS
PACIFIC-2	Unresectable stage III NSCLC	Concurrent durvalumab strategy was negative
LAURA	EGFR+ unresectable stage III NSCLC	Osimertinib after CRT transformed outcomes
INT 0139	Stage IIIA(N2) resectable	No OS benefit for routine trimodality; pneumonectomy problematic
EORTC 08941	Stage IIIA(N2) responders	Surgery was not superior to RT after induction chemotherapy
CheckMate 816	Resectable NSCLC	Neoadjuvant nivo + chemo dramatically improved pCR
KEYNOTE-671, AEGEAN	Resectable NSCLC	Perioperative IO improved event-free outcomes
Lung ART	Completely resected pN2 NSCLC	Modern PORT did not improve DFS or OS
Intergroup 0096 / Turrisi	LS-SCLC	45 Gy BID improved survival vs daily 45 Gy
CONVERT	LS-SCLC	66/33 daily was equivalent to BID treatment
CALGB 30610	LS-SCLC	70/35 daily also performed similarly to BID
ADRIATIC	LS-SCLC after CRT	Adjuvant durvalumab became the new standard
NRG-LU005	LS-SCLC	Concurrent atezolizumab + CRT was negative
Slotman thoracic RT (2015)	ES-SCLC responders	Supported selective consolidative thoracic RT
Auperin	LS-SCLC CR	PCI improved survival by about 5%
Le Pechoux	LS-SCLC PCI dose	25 Gy remained the PCI standard
Slotman EORTC (2007)	ES-SCLC responders	PCI improved brain control and OS before routine MRI surveillance
Takahashi (2017)	ES-SCLC with MRI surveillance	PCI did not improve OS in the MRI era
GOECP-SEOR	HA-PCI	Positive randomized signal for cognitive preservation
NRG-CC003	HA-PCI	Mixed result, but composite neurocognitive outcomes favored HA-PCI