- Takeaway:
- Concomitant chemotherapy + radiation therapy (RT):
- Yields a statistically significant OS benefit vs RT alone:
- HR ~ 0.83; ~ 6.5% 5-yr OS gain:
- Driven by improved locoregional control
- HR ~ 0.83; ~ 6.5% 5-yr OS gain:
- Induction alone or adjuvant alone did not improve OS
- Yields a statistically significant OS benefit vs RT alone:
- Concomitant chemotherapy + radiation therapy (RT):
- Use:
- This underpins cisplatin-CRT as the default curative standard
- Refs:
- Lacas et al., MACH-NC (Radiother Oncol 2021; PMC)
- What it is individual-patient data meta-analysis updating the MACH-NC database:
- To 107 RCTs / 19,805 patients:
- Accrued 1965 to 2012
- Updated median follow-up:
- 6.6 years
- Endpoints:
- OS, EFS, 120-day mortality, loco-regional failure (LRF), distant failure (DF), cancer vs non-cancer deaths Groningen Research Portal+1
- Results by timing of chemotherapy:
- Concomitant chemoradiotherapy (CRT) vs loco-regional therapy alone:
- Overall survival:
- HR 0.83 (95% CI 0.79–0.86):
- Absolute OS gain + 6.5% at 5 yr and + 3.6% at 10 yr (≈ NNT ~ 15 at 5 yr)
- HR 0.83 (95% CI 0.79–0.86):
- EFS:
- HR 0.80 with + 5.8% at 5 yr
- No increase in 120-day mortality:
- HR 1.07, p=0.37 Groningen Research Portal
- Patterns of failure:
- Marked LRF reduction (sub-HR 0.71)
- No DF effect (HR 1.04)
- Survival benefit driven by lower cancer mortality (HR 0.79; − 9.8% at 5 yr):
- With no change in non-cancer deaths (HR 1.01) Groningen Research Portal
- Regimen effects:
- Greatest effect for platinum-containing polychemotherapy (EFS HR 0.74)
- Smallest for non-platinum monochemotherapy (HR 0.86)
- Benefit consistent across eras and RT modalities Groningen Research Portal
- Overall survival:
- Concomitant chemoradiotherapy (CRT) vs loco-regional therapy alone:
- Induction chemotherapy (before RT / CRT):
- OS and EFS:
- No significant benefit:
- OS HR 0.96
- EFS HR 0.96
- DF decreases (HR 0.76):
- But no LRF improvement (sub-HR 1.07)
- Benefit attenuates with worse performance status Groningen Research Portal
- No significant benefit:
- OS and EFS:
- Adjuvant chemotherapy (after loco-regional therapy):
- OS and EFS:
- No benefit:
- OS HR 1.02
- EFS HR 0.98
- No benefit:
- Early mortality:
- Higher 120-day mortality (HR 1.89, p=0.0003)
- Some reduction in LRF (sub-HR 0.84) and DF(HR 0.77) without survival gain Groningen Research Portal
- OS and EFS:
- Direct head-to-head concomitant vs induction (same drugs):
- Across eight comparisons (n=1,214), concomitant is superior:
- OS HR 0.84 (≈ + 6.2% absolute OS at 5 yr)
- EFS HR 0.85
- LRF HR 0.86
- Indirect comparisons concur (interaction p<0.0001) Groningen Research Portal
- Across eight comparisons (n=1,214), concomitant is superior:
- Subgroups and modifiers:
- Age:
- Benefit of concomitant chemotherapy decreases with increasing age:
- Use added caution > 70 yr Groningen Research Portal
- Benefit of concomitant chemotherapy decreases with increasing age:
- Performance status:
- Poorer PS attenuates benefit:
- Noted in induction analyses:
- Similar theme in concomitant Groningen Research Portal
- Noted in induction analyses:
- Poorer PS attenuates benefit:
- HPV / smoking:
- Not evaluable here (most trials pre-HPV era)
- Don’t expect HPV-specific signals from MACH-NC Groningen Research Portal
- Age:
- Practical takeaways:
- For curative, non-metastatic HNSCC:
- The only timing with proven OS benefit is concomitant chemo with RT
- Expect ~ 6% to 7% absolute 5-yr OS gain:
- Driven by better loco-regional control Groningen Research Portal
- Induction doesn’t improve OS vs loco-regional therapy alone (despite fewer distant mets)
- Adjuvant chemo doesn’t improve OS and raises early mortality risk Groningen Research Portal+1
- Within concomitant therapy:
- Platinum-containing regimens carry the strongest signal:
- This underpins cisplatin-based CRT as the benchmark Groningen Research Portal
- Platinum-containing regimens carry the strongest signal:
- For curative, non-metastatic HNSCC:
- To 107 RCTs / 19,805 patients:
- Limitations the paper notes:
- Trials span 1965 to 2012:
- Staging, RT techniques, supportive care evolved
- Still, no interaction by accrual period and sensitivity analyses were consistent
- HPV data were largely unavailable. Groningen Research Portal
- Trials span 1965 to 2012:
- Citation (free full text):
- Lacas B, Carmel A, Landais C, et al. MACH-NC update—107 RCTs, 19,805 pts. Radiother Oncol. 2021;156:281-293. PMC8386522. Key effect sizes and table are in the PDF’s Table 1 and text (OS/EFS/120-day mortality/LRF/DF).
Rodrigo Arrangoiz MS, MD, FACS, FSSO 