Head and Neck Cancer — Board Review Summary
PART I — HPV+ OROPHARYNX
TORS vs Primary RT/CRT
| Trial | Design | Key results | Board takeaway |
|---|---|---|---|
| ORATOR | Phase II RCT, N=68, T1-2 N0-2 (≤4 cm) OPSCC, HPV+/-; RT 70 Gy +/- chemo vs TORS + neck dissection +/- adjuvant therapy | MDADI at 1y favored RT: 86.9 vs 80.1 (p=0.042), but difference did not meet the prespecified 10-point threshold; 5y MDADI converged; OS/PFS excellent in both arms | No survival-powered head-to-head winner; toxicity spectra differ more than efficacy |
| ORATOR2 | Phase II RCT, N=61, HPV+ only; de-escalated RT 60 Gy +/- weekly cisplatin vs de-escalated TOS + ND +/- reduced adjuvant RT | Stopped early for excess mortality in surgical arm; 2y OS 100% RT vs 90% TOS; 2y PFS 100% vs 86% | De-escalated surgery was not safe in this trial; de-escalated primary RT performed extremely well |
High-yield TORS principle: choose the modality that uses the fewest treatment modalities necessary. Avoid triple-modality treatment whenever possible. TORS fits best for well-lateralized, well-exposed T1-2 tonsil/base-of-tongue tumors with low ipsilateral nodal burden; bulky N2-3 disease, obvious ENE, poor exposure, or infiltrative midline disease favor primary CRT.
Definitive RT Alone vs CRT
| Clinical setting | Preferred approach | Supporting detail |
|---|---|---|
| T1-2 N0 | RT alone reasonable | Strong ASTRO support; excellent outcomes in carefully selected early-stage disease |
| T1-2 N1, single node ≤3 cm | RT alone acceptable in selected HPV+ cases | Especially when chemo is contraindicated |
| T3-4 any N | CRT | Locally advanced disease standard |
| ≥2 nodes or single node >3 cm | CRT | ASTRO / NCCN standard |
RTOG 1016 established that cetuximab is inferior to cisplatin in HPV+ OPC. With accelerated IMRT 70 Gy / 35 fx / 6 wk, 5y OS was 84.6% with cisplatin vs 77.9% with cetuximab; 5y PFS 78.4% vs 67.3%; locoregional failure 9.9% vs 17.3%. Severe toxicity rates were similar. Cisplatin + RT remains standard; cetuximab is only for true cisplatin-ineligible patients.
Definitive Dose / Fractionation / Volumes
| Scenario | Dose / fractionation | Notes |
|---|---|---|
| Gross disease with concurrent systemic therapy | 70 Gy / 33-35 fx | ASTRO strong recommendation |
| Elective neck, sequential | 44-50 Gy | Conventional elective range |
| Elective neck, SIB | 54-63 Gy | Typically 1.6-1.8 Gy/fx |
| RT alone, conventional | 68-70 Gy | T1-2 N0-1 single node ≤3 cm |
| RT alone, hypofractionated | 66-70 Gy at 2.1-2.2 Gy/fx | Fewer visits |
| RT alone, accelerated | 68-72 Gy over 6 weeks | Shorter overall time |
| Hyperfractionated RT | 74.4-81.6 Gy, 1.2 Gy BID | Strongly recommended for T3-4 when treating RT-alone style regimens |
Elective Nodal Selection and Unilateral RT
Common elective levels: II-IV, retropharyngeal, and retrostyloid depending on nodal status.
Often omissible: level IB if no oral cavity involvement; level V if no naso/hypopharyngeal involvement.
Unilateral RT strong recommendation: HPV+ T1-2 palatine tonsil cancer confined to tonsillar fossa with either no positive node or a single node ≤3 cm and no ENE.
Bilateral RT required: base of tongue, soft palate, posterior pharyngeal wall, or tonsil tumors crossing midline / invading tongue base.
Often omissible: level IB if no oral cavity involvement; level V if no naso/hypopharyngeal involvement.
Unilateral RT strong recommendation: HPV+ T1-2 palatine tonsil cancer confined to tonsillar fossa with either no positive node or a single node ≤3 cm and no ENE.
Bilateral RT required: base of tongue, soft palate, posterior pharyngeal wall, or tonsil tumors crossing midline / invading tongue base.
Definitive De-escalation: What Failed vs What May Still Work
NRG-HN005 was the sobering reminder that blanket dose reduction fails: 60 Gy + cisplatin and 60 Gy + nivolumab did not demonstrate non-inferiority versus standard 70 Gy + cisplatin. Clinical selection by HPV status and smoking history alone is not enough.
Investigational, not standard: biomarker-driven de-escalation is still alive. The MSK 30 ROC hypoxia-guided pilot used FMISO PET to select candidates for 30 Gy / 15 fx + platinum, with 2y locoregional control about 94.4% and no grade 3 RT-related mucositis/dysphagia. Other response-adapted strategies include mid-RT FDG-PET, mid-RT CT nodal response, and ctHPV DNA kinetics, but these remain trial-oriented.
Postoperative De-escalation in HPV+ OPC
| Study | Key design | Main result | Takeaway |
|---|---|---|---|
| ECOG 3311 | After TORS, low-risk observed; intermediate-risk randomized to 50 Gy vs 60 Gy; high-risk got 66 Gy + weekly cisplatin | 54-mo PFS: 94.9% for 50 Gy, 90.2% for 60 Gy; low-risk observation also excellent | 50 Gy is a viable intermediate-risk postoperative dose in selected HPV+ patients |
| MC1675 / DART | Intermediate-risk without ENE: de-escalated 30 Gy BID + weekly docetaxel vs 60 Gy; high-risk ENE cohort also randomized | Intermediate-risk cohort: excellent disease control with lower toxicity / better QOL; high-risk ENE cohort had worse PFS with de-escalation | Very aggressive de-escalation may work in favorable intermediate-risk disease, but be cautious in ENE-positive / pN2 disease |
Mucosal-sparing postoperative RT in select resected p16+ OPC treated only the nodal regions and omitted the primary mucosal site. At 2 years: 98% local control, 97% locoregional control, 100% OS, with only 3.3% PEG use overall and 0 PEG tubes in the proton cohort. Very promising, but still a selected strategy rather than universal standard.
Post-op HPV+ dose guide: high-risk ENE and/or positive margins 60-66 Gy; intermediate-risk 56-60 Gy; ECOG 3311-style de-escalation to 50 Gy is reasonable in carefully selected intermediate-risk HPV+ patients.
PART II — NASOPHARYNX CANCER
RT vs CRT vs IC-CRT
| Trial | Design | Key result | Takeaway |
|---|---|---|---|
| Intergroup 0099 | Stage III-IV NPC; RT alone vs RT 70 Gy + concurrent cisplatin then adjuvant PF | 3y PFS 69% vs 24%; 3y OS 78% vs 47% | Historic landmark establishing CRT-based intensification |
| Chen 2012 | CCRT + adjuvant PF vs CCRT alone | No significant benefit in FFS or OS | Adjuvant PF after modern CCRT is controversial |
| NPC-0502 | Post-CRT EBV-positive patients; adjuvant GP vs observation | No significant benefit | Biomarker-selected adjuvant PF/GP has not clearly rescued AC |
| Metronomic capecitabine | High-risk LA-NPC after CCRT +/- IC; capecitabine 650 mg/m2 BID x 1 year vs observation | 3y FFS 85.3% vs 75.7%; OS 93.3% vs 86.1%; DMFS 89.2% vs 79.7% | One of the strongest adjuvant-era positive studies |
| TPF IC | TPF x3 then CCRT vs CCRT alone | 5y OS 85.6% vs 77.7% | IC improves distant control and survival |
| GP IC | Gemcitabine/cisplatin x3 then CCRT vs CCRT alone | Final 5y OS 87.9% vs 78.8% | GP is a modern preferred IC regimen |
| MAC-NPC | Network meta-analysis, 8,221 pts / 31 RCTs | IC + CCRT ranked highest for OS, PFS, distant progression, and NPC-specific death | IC + CCRT is the preferred overall strategy in LA-NPC |
| Guo 2025 | TPF IC + CCRT vs CCRT + AC in N2-3, EBV-high-risk NPC | No PFS winner; IC better DMFS, AC better LR-FFS | IC and AC solve different problems: distant vs local-regional control |
NCCN 2026 practical framework: T1N0 and many low-risk T2N0 cases can still be treated with RT alone; CRT is the backbone for intermediate disease; and IC -> CRT is preferred for locoregionally advanced disease, especially bulky / EBV-high-risk presentations.
Elective Nodal Coverage: UNI vs WNI
Tang et al. phase III randomized 446 N0-N1 NPC patients to unilateral-neck irradiation (UNI) vs bilateral whole-neck irradiation (WNI). UNI was non-inferior: 5y regional RFS 95.0% vs 94.9%, OS 95.9% vs 93.1%. Only 2/446 patients had contralateral lower-neck recurrence, one in each arm.
| Late toxicity in 5y disease-free survivors | WNI | UNI |
|---|---|---|
| Hypothyroidism | 48% | 34% |
| Neck tissue damage / fibrosis | 46% | 29% |
| Dysphagia | 27% | 14% |
| Carotid stenosis | 26% | 15% |
NPC nodal-coverage pearl: uninvolved contralateral lower neck is at extremely low risk in properly selected N0-N1 disease. Modern NPC practice is moving away from routine whole-neck treatment for everyone.
Post-Induction RT Dose / Volume De-escalation
| Study | Strategy | Main result |
|---|---|---|
| Yang / Xiang phase III | Post-IC residual GTV got 70 Gy in both arms; resolved pre-IC GTV randomized to 70 Gy vs 64 Gy | No difference in disease control or survival; 64 Gy arm had better xerostomia / cognitive outcomes |
| Tang 2025 phase III | Post-IC soft-tissue GTV volume reduction vs conventional pre-IC GTV; both arms 70 Gy to GTV, 60-62 Gy high-risk CTV, 54-56 Gy low-risk CTV | 3y LRRFS 91.5% vs 91.2%, non-inferior; less acute mucositis and less late otitis/xerostomia with reduced-volume planning |
NPC de-escalation rule to memorize: post-IC soft-tissue response can guide radical-dose volume reduction, but bony structures with initial invasion should still be delineated from pre-IC imaging and treated aggressively. IG-2024 moved toward endorsing post-IC GTV for the 70 Gy equivalent volume and lowered elective dose to about 50 Gy equivalent.
PART III — EARLY GLOTTIC LARYNX
T1N0 Partial Larynx / Single Vocal Cord RT
| Study | Population | Local control | Functional / toxicity signal |
|---|---|---|---|
| Erasmus MC / Tans 2022 | T1a mostly | 5y LC 97.1% | Late grade 3+ 6.5%; voice improved over time |
| Al-Mamgani 2015 | T1a N0 | 2y LC 100% vs 92.2% WLRT | Acute toxicity 17% vs 66% with WLRT; better voice outcomes |
| Elsharief 2026 RCT | T1a N0 | 2y LC 100% SVCI vs 96% WLRT | Better VHI at 1y; no grade 3-4 toxicity |
| MD Anderson 2025 | T1-2/Tis N0 | 3y LRC 85.4% PLRT vs 90.8% WLRT | Less acute dysphagia and better MDADI with PLRT |
Main point: for T1a N0, partial-larynx or single-vocal-cord RT can achieve 97-100% local control with better voice / acute-toxicity profiles. For T2, data are less secure and local control may be lower, so caution is warranted.
NCCN Dose Standards for Early Glottic RT
| Stage | Dose options |
|---|---|
| T1 N0 | 63 Gy / 28 at 2.25 Gy/fx preferred; or 66 Gy / 33; or 50-52 Gy at 3.12-3.28 Gy/fx |
| T2 N0 | 64.8 Gy at 2.4 Gy/fx up to 70 Gy / 35 |
T2N0 Glottic: Add Chemo or Intensify RT?
T2N0 glottic disease has worse local control with RT alone than T1 disease, especially with impaired cord mobility, subglottic extension, anterior commissure involvement, or bulky GTV. Retrospective series suggest CRT can improve local control and voice preservation, but there is no dedicated randomized phase III CRT-vs-RT trial for T2N0 glottic cancer.
Practical board takeaway: "unfavorable T2" glottic cancer may justify treatment intensification either through CRT or altered-fractionation / accelerated RT. The deck's institutional examples favor considering chemo when there is possible paraglottic involvement, bulky primary, subglottic extension, or sluggish cords.
PART IV — LARYNX / HYPOPHARYNX, HPV-NEGATIVE: ELECTIVE NODAL DE-ESCALATION
Why This Is Different from Failed Gross-Disease De-escalation
The lecture makes a key conceptual distinction: gross-disease de-escalation failed in HPV+ OPC because macroscopic tumor has a steep dose-response curve. Elective nodal de-escalation targets microscopic disease, where the dose-response curve is shallower and de-escalation is therefore biologically more plausible.
Randomized and Prospective Data
| Study | Population / design | Elective dose | Main result |
|---|---|---|---|
| Nevens / Deschuymer | Belgian phase III RCT, N=200, ~90% HPV-negative | 40 Gy vs 50 Gy | Only 2 true elective failures per arm; lower salivary toxicity with 40 Gy |
| UPGRADE-RT | Dutch phase III RCT, N=300, RT alone for OP/larynx/hypopharynx | 43 Gy vs 50 Gy | 2y elective nodal recurrence 4.9% vs 4.3%; less acute grade 3+ dysphagia with 43 Gy |
| Sher phase II | Dual-center, mixed HPV, most with concurrent chemo | 40 Gy / 20 + reduced elective volumes | 0 solitary elective nodal failures |
| Zakeri 2025 | 73 consecutive HPV-negative larynx / hypopharynx / p16-neg OPC / CUP pts | 40 Gy elective with CRT | 0 solitary elective nodal recurrences; all locoregional recurrences were in 70 Gy volumes |
| DEPEND | HPV-negative stage IVa/b after neoadjuvant nivo + carbo/taxol; deep responders omitted elective volumes | 0 Gy elective in responders | 2y PFS 66%, OS 73%, with less mucositis, dermatitis, and dry mouth in de-escalated responders |
Guideline Position
NCCN 2026 still lists elective dose as 44-50 Gy sequential or 54-63 Gy SIB. ASCO SCCUP 2020 already endorses a biologically equivalent elective dose range of roughly 40-50 Gy to negative nodal regions. So 40 Gy is evidence-based, but has been incorporated into guidelines unevenly.
PART V — ORAL CAVITY: PORT, CRT, AND PERIOPERATIVE IMMUNOTHERAPY
Surgery First, Then Pathology-Driven Adjuvant Therapy
| Risk group | Features | Usual treatment implication |
|---|---|---|
| High risk | ENE and/or positive margin | Category 1 postoperative CRT |
| Intermediate risk | Close margins, pT3-4, pN2-3, level IV/V nodes, PNI, vascular invasion, lymphatic invasion | PORT, with selective discussion of systemic intensification |
Landmark Postoperative Trials
| Trial | Design | Main result | Board takeaway |
|---|---|---|---|
| EORTC 22931 | Post-op RT 66 Gy vs CRT with cisplatin 100 mg/m2 q3wk x3 | 5y OS 53% vs 40%; improved PFS and locoregional control | Post-op CRT clearly better in very high-risk patients |
| RTOG 9501 | Post-op RT 60 Gy +/- cisplatin | Long-term benefit confined mainly to ENE / positive margin subgroup | ≥2 nodes alone did not clearly benefit from cisplatin |
| Bernier pooled 2005 | Pooled EORTC + RTOG | ENE and/or positive margins were the only features with consistent CRT benefit in both trials | This created the modern ENE/+margin paradigm |
| Zumsteg 2025 | Updated combined analysis, n=744 | CRT improved OS overall (HR 0.81) and reduced cancer-specific mortality, but increased other-cause mortality; no interaction showing ENE/+margin uniquely predictive | Provocative challenge to the old paradigm, but does not erase the current guideline standard |
Current exam answer: ENE and positive margins remain the strongest category 1 indications for postoperative CRT. The 2025 updated pooled analysis is important and provocative, but it should be interpreted as a challenge to dogma rather than a wholesale guideline replacement.
Postoperative Systemic Options
- Preferred: cisplatin 100 mg/m2 on days 1, 22, and 43 of RT.
- Common alternate schedule: weekly cisplatin 40 mg/m2.
- Cisplatin-ineligible: docetaxel, or docetaxel + cetuximab in selected cases, but with weaker evidence.
Tian JCO 2024 — Early Oral Tongue pT1-2N0M0
In propensity-matched early oral tongue SCC, PORT improved outcomes in several adverse-feature subgroups. The strongest signal was for moderate-to-poor differentiation, where OS was 97% vs 69% and DFS 88% vs 50%. PNI/LVI, DOI >5 mm, and close margins also showed benefit. This is not yet a universal guideline rewrite, but it is a very testable modern nuance.
KEYNOTE-689
KEYNOTE-689 randomized 714 patients with resectable stage III-IVA HNSCC to perioperative pembrolizumab vs standard care. At 36 months, EFS favored pembrolizumab across CPS-defined groups: total population 57.6% vs 46.4% (HR 0.73), CPS ≥1 58.2% vs 44.9%, CPS ≥10 59.8% vs 45.9%. Surgery completion was about 88% in both arms. NCCN 2026 incorporates perioperative pembrolizumab for resectable PD-L1-positive locally advanced disease.
Post-op oral cavity dose pearl: high-risk volumes generally receive 60-66 Gy, with RT ideally beginning within 6 weeks of surgery.
CROSS-CUTTING HIGH-YIELD POINTS
- ORS / TORS vs RT in HPV+ OPC: only ORATOR and ORATOR2 directly randomized surgery vs RT/CRT; there is still no survival-powered phase III head-to-head trial.
- ORATOR: RT and TORS both yielded excellent oncologic outcomes; swallowing favored RT at 1 year, then converged.
- ORATOR2: stopped early because of excess surgical-arm mortality; de-escalated RT performed extremely well.
- RTOG 1016: cetuximab is inferior to cisplatin in HPV+ OPC and should not replace cisplatin in eligible patients.
- HPV+ definitive standard: gross disease 70 Gy / 33-35 fx with cisplatin.
- RT alone in HPV+ OPC: acceptable mainly for T1-2 N0-1 with a single node ≤3 cm or when chemo is contraindicated.
- NRG-HN005: unselected dose reduction to 60 Gy failed; blanket definitive de-escalation is not standard.
- Unilateral RT in HPV+ tonsil: reasonable when well lateralized and low nodal burden; BOT / soft palate / PPW / midline disease requires bilateral treatment.
- ECOG 3311: postoperative 50 Gy is a key intermediate-risk HPV+ dose after TORS.
- MC1675: aggressive postoperative de-escalation may work in intermediate-risk disease, but ENE-positive patients did worse with overly reduced therapy.
- NPC systemic backbone: CCRT remains essential, but modern data favor IC + CCRT over CCRT + AC overall.
- NPC adjuvant nuance: classic PF after CCRT is controversial, but metronomic capecitabine is a real positive adjuvant study.
- NPC nodal coverage: unilateral-neck approaches can safely spare the contralateral lower neck in selected N0-1 patients and reduce late toxicity.
- Post-IC NPC planning: residual soft-tissue GTV can guide the 70 Gy volume, but initially involved bone should still be respected from pre-IC imaging.
- T1a glottic cancer: partial-larynx / single-vocal-cord RT is increasingly compelling, with excellent control and better voice / toxicity outcomes.
- T2 glottic cancer: worse local control than T1; consider intensified RT or CRT for unfavorable features.
- HPV-negative larynx/hypopharynx elective de-escalation: randomized data now support 40-43 Gy as a safe elective dose in selected settings.
- Oral cavity surgery-first principle: ENE and positive margins remain the strongest indications for postoperative CRT.
- Zumsteg 2025: important challenge to the ENE/+margin-only dogma, but not yet a wholesale guideline reversal.
- KEYNOTE-689: first major perioperative immunotherapy advance incorporated into oral cavity / resectable LA HNSCC pathways.
CONSOLIDATED DOSE TABLE
| Setting | Dose / regimen | Why it matters |
|---|---|---|
| HPV+ OPC definitive CRT | 70 Gy / 33-35 fx | Modern standard gross-disease dose |
| HPV+ OPC elective neck, sequential | 44-50 Gy | Common conventional elective range |
| HPV+ OPC elective neck, SIB | 54-63 Gy | SIB elective range |
| HPV+ OPC RT alone | 68-70 Gy | Early-stage / chemo-ineligible approach |
| 30 ROC pilot | 30 Gy / 15 fx | Investigational hypoxia-guided major de-escalation |
| ECOG 3311 intermediate risk | 50 Gy | Key postoperative HPV+ de-escalation dose |
| Post-op HPV+ high risk | 60-66 Gy | Positive margin / ENE postoperative range |
| NPC RT alone, early stage | 70 Gy | Backbone RT-alone dose |
| NPC post-IC resolved pre-IC GTV | 64 Gy | Phase III-supported reduced dose to resolved pre-IC disease |
| NPC high-risk CTV | 60-62 Gy | Tang reduced-volume phase III schema |
| NPC low-risk CTV | 54-56 Gy | Conventional low-risk dose in Tang phase III |
| Early glottic T1 N0 | 63 Gy / 28 | NCCN preferred conventional regimen |
| Early glottic T1 N0 alternative | 50-52 Gy | Hypofractionated option |
| Glottic T2 N0 | 64.8-70 Gy | Standard definitive RT range |
| HPV-negative elective nodal de-escalation | 40-43 Gy | Supported by Nevens / UPGRADE-RT |
| Oral cavity post-op RT | 60-66 Gy | High-risk postoperative range |
| Post-op concurrent cisplatin | 100 mg/m2 | Days 1, 22, 43 standard schedule |
KEY LANDMARK TRIALS (memorize)
| Trial | Disease | One-line takeaway |
|---|---|---|
| ORATOR | HPV+/- OPSCC | TORS and RT both work; functional tradeoffs differ more than control outcomes |
| ORATOR2 | HPV+ OPSCC | De-escalated surgery arm had excess mortality; de-escalated RT was strong |
| RTOG 1016 | HPV+ OPC | Cetuximab is inferior to cisplatin with RT |
| NRG-HN005 | HPV+ OPC | Blanket definitive dose de-escalation to 60 Gy failed |
| ECOG 3311 | Post-op HPV+ OPC | 50 Gy is a key intermediate-risk postoperative dose after TORS |
| MC1675 / DART | Post-op HPV+ OPC | Aggressive de-escalation may work in favorable intermediate risk, but not safely in higher-risk ENE disease |
| Mucosal-sparing RT | Resected HPV+ OPC | Regional-only postoperative RT is promising in selected patients |
| Intergroup 0099 | NPC | Established CRT-based intensification as the landmark standard |
| Chen 2012 | NPC | Adjuvant PF after CCRT did not clearly improve outcomes |
| Chen 2021 capecitabine | NPC | Metronomic adjuvant capecitabine improved FFS, OS, and DMFS |
| TPF (Sun 2016) | NPC | Induction TPF improved long-term survival over CCRT alone |
| GP (Zhang 2019) | NPC | Induction gemcitabine/cisplatin is a modern preferred IC regimen |
| MAC-NPC | NPC | IC + CCRT ranks highest overall among combined-modality strategies |
| Tang UNI vs WNI | NPC nodal coverage | Upper-neck-only sparing is non-inferior in selected N0-1 patients and reduces late toxicity |
| Yang / Xiang | NPC post-IC planning | Resolved pre-IC disease can safely receive intermediate rather than full dose |
| Tang 2025 reduced-volume RT | NPC post-IC planning | Post-IC reduced-volume RT is non-inferior and less toxic |
| UPGRADE-RT | HPV-negative HN elective neck | 43 Gy elective dose is safe and less toxic than 50 Gy |
| Nevens / Deschuymer | HPV-negative HN elective neck | 40 Gy elective dose produced very few true elective failures |
| EORTC 22931 | Post-op oral cavity / HNSCC | CRT improved OS, PFS, and locoregional control vs RT alone |
| RTOG 9501 | Post-op oral cavity / HNSCC | Long-term CRT benefit was strongest in ENE / positive-margin disease |
| Bernier pooled 2005 | Post-op HNSCC | Created the ENE/+margin-only paradigm for category 1 postoperative CRT |
| Zumsteg 2025 | Post-op HNSCC | Challenges the strict ENE/+margin-only paradigm, but with toxicity tradeoffs |
| KEYNOTE-689 | Resectable locally advanced HNSCC | Perioperative pembrolizumab improved EFS and is now NCCN-incorporated |
| Tian 2024 | Early oral tongue SCC | PORT may help selected pT1-2N0 patients with adverse features, especially moderate-poor differentiation |