Intra-arterial vs. Intravenous Chemoradiation for Advanced Head and Neck Cancer, Early Results of a Multi-institutional Trial

Reviewer: John P. Plastaras, MD, PhD
Abramson Cancer Center of the University of Pennsylvania
Last Modified: November 7, 2006

Presenter: C. R. N. Rasch, M.D., Ph.D.
Presenter's Affiliation: Netherlands
Type of Session: Plenary


  • The addition of concurrent cisplatin-based chemotherapy to definitive radiotherapy for locally advanced squamous cell carcinoma of the head and neck (HNSCC) improves local control and survival.
  • Cisplatin has significant systemic side-effects, including nephrotoxicity, ototoxicity, and bone marrow suppression, which ultimately limit the maximum dose to the tumor.
  • Intra-arterial (via the carotid or lingual artery) cisplatin administered with a systemic scavenger allows doses 5-10 times higher, leading to higher levels of tumor cisplatin adducts.
  • The authors hypothesized that higher cisplatin levels delivered intra-arterially, concurrently with radiation, can improve locoregional control compared to intravenous delivery.

Materials and Methods

  • Phase III, multicenter trial. Patients were randomized to:
    • Arm A: Intravenous ? 100 mg/m2 x 3 (days 1, 22, and 43)
    • Arm B: Intra-arterial ? 150 mg/m2 x 4 (days 2, 9, 16, and 23) with systemic rescue (sodium thiosulfate)
  • All patients underwent radiotherapy: 70 Gy (2 Gy x 35 fractions)
  • Patients:
    • 240 HNSCC patients, 240 analyzed
    • Inclusion: T3/T4 inoperable or functionally inoperable, Stage IV, oropharynx/oral cavity/hypopharynx
    • Breakdown by stages: 3% T2, 31% T3, 67% T4
  • Endpoints:
    • Primary: Local-regional control.
    • Powered to detect a 15% difference


  • Compliance was good (78% in intra-arterial, 80% intravenous)
  • Complete response was better in intravenous: 68% intra-arterial, 81% intravenous, p=0.02
  • There was no significant difference in 2-yr local-regional control: 62% intra-arterial, 68% intravenous
  • There was no significant difference in 2-yr distant metastases rate or 2-yr overall survival (61% intra-arterial, 63% intravenous)
  • Toxicity (> Grade 2): There were fewer renal events but more neurologic events in the intra-arterial arm:
    • Renal: 1 intra-arterial, 9 intravenous (p=0.02)
    • Hematologic: 52 intra-arterial, 42 intravenous
    • Ototoxicity: 53 intra-arterial, 58 intravenous (p=0.395)
    • The biggest improvement was observed in patients whose hearing was good prior to treatment.
    • Skin: 10 intra-arterial, 23 intravenous
    • Cardiac/pneumonia: 5 intra-arterial, 9 intravenous
    • Neurologic; 9 intra-arterial (including 1 CVA), 1 intravenous, p=0.005
  • Large tumor size predicted for worse local-regional control.

Author's Conclusions

  • Local-regional control was not different in HNSCC patients treated definitively with concurrent chemoradiotherapy whether the chemotherapy was given intravenously or intra-arterially
  • Large tumor size predicted for poorer outcomes.
  • Intra-arterial cisplatin was associated with less nephrotoxicity, somewhat less ototoxicity, but more neurologic toxicity as compared to intravenous administration.

Clinical/Scientific Implications

According to the discussant, Dr. K. Thomas Robbins, one possible reason for this negative result may be the fact that the Dutch group used more bilateral intraarterial infusions, as compared to what is generally done in the U.S. This bilateral infusion technique was used even when the primary HNSCC lesions extended only slightly across midline. As a result, there may have been some under-dosing of the tumor because of the divided chemotherapy administration.