Background
Phase III randomized trial (RTOG 83-05) comparing two RT fractionation schedules for unresectable or metastatic melanoma. 137 patients. Conventional RT: 2.5 Gy/fx × 20 fractions = 50 Gy. Hypofractionated RT: 8 Gy/fx × 4 fractions = 32 Gy, one fraction per week. Primary endpoint: complete response rate.
Interventions and follow up
Arm A: Hypofractionated RT — 8 Gy × 4 fractions (32 Gy total), once weekly
Arm B: Conventional RT — 2.5 Gy × 20 fractions (50 Gy total)
Primary endpoint: Complete response rate
mFollow up: Not specified
Arm B: Conventional RT — 2.5 Gy × 20 fractions (50 Gy total)
Primary endpoint: Complete response rate
mFollow up: Not specified
Results
Complete Response (hypofractionation): 24%
Complete Response (conventional): 35%
Difference: Not statistically significant
Response duration: Similar between arms; no significant difference in OS or DFS
Complete Response (conventional): 35%
Difference: Not statistically significant
Response duration: Similar between arms; no significant difference in OS or DFS
Adverse events
Main adverse events: Hypofractionation (8 Gy/fx): higher acute skin reaction during and immediately post-RT; late toxicity also higher with 8 Gy/fx at some sites. Conventional fractionation: better tolerability for treatment near critical structures. No unexpected grade 4 toxicities in either arm.
Conclusions
Both hypofractionated and conventional RT schedules achieve similar complete response rates (~24–35%) for melanoma, with no statistically significant difference. This trial established that hypofractionation is an acceptable RT approach for melanoma, though 8 Gy × 4 may have modestly lower response than conventional fractionation.
Key Limitations
Key Limitations: Small sample; heterogeneous tumor sites and histologic subtypes within "melanoma" label; pre-imaging era assessment of response; 8 Gy/fx schedule is more toxic than the 6 Gy/fx currently used at most centers; modern SBRT (12–20 Gy single fraction) not evaluated.
Clinical Context
Historical landmark establishing that melanoma is NOT uniquely radioresistant to hypofractionation, contrary to early dogma. Combined with MDACC series (Ballo 2003, Chang 2006), supports use of 6 Gy × 5 = 30 Gy as the practical standard for melanoma nodal RT. The 8 Gy/fx arm had slightly lower CR rate, leading to preference for 6 Gy/fx over 8 Gy/fx in modern practice.