Nonrandomized trials have suggested that intra-arterial delivery of chemoradiation would be superior to intravenous delivery of comparable chemoradiation in patients with inoperable head and neck cancer. Now, a randomized multi-institutional phase III trial demonstrates that there is no difference in clinical outcomes for patients treated with one method versus the other. Not everyone, however, is convinced that it is time to move to other approaches.
Surprisingly, there were no differences, no differences at all, in terms of tumor control between the two arms of the trial, said Coen Rasch, MD, Radiation Oncologist at the Netherlands Cancer Institute in Amsterdam, who presented the data at the annual meeting of the American Society for Therapeutic Radiology and Oncology (ASTRO) last fall. Since the current standard in the community is intravenous [delivery], we think intravenous will remain the standard of care.
Dr. Rasch’s team randomly assigned 240 patients with inoperable head and neck cancer to one of two arms: four courses of 150 mg/m2 cisplatin delivered intra-arterially into a carotid artery and intravenous sodium thiosulfate on days 2, 9, 16, and 23 (IA arm) or 100 mg/m2 intravenous cisplatin on days 1, 22, and 43 (IV arm). All patients received concurrent radiation of 70 Gy, delivered in seven fractions. Dr. Rasch and colleagues found that locoregional disease was controlled in 62% of patients in the IA arm and 68% of those in the IV arm at two years. Overall survival at two years was 61% and 63% in the IA and IV arms, respectively.
Theodore S. Lawrence, MD, PhD, Professor and Chair of the Department of Radiation Oncology at the University of Michigan Hospitals and Health System in Ann Arbor, and immediate past president of ASTRO, who led a press conference in which the data were presented, was philosophical about the results. Sometimes a very important study like this is a negative study, he said. This was a technique that is very difficult to do but very promising in the early results. Now we know, through this carefully carried-out study, that we should focus our energies on new approaches elsewhere. Sometimes that is a very important finding.
Abandon IA? Not Yet
Not everyone is ready to abandon the IA approach, however.
K. Thomas Robbins, MD, Director of the SimmonsCooper Cancer Institute and Professor of Otolaryngology-Head and Neck Surgery at Southern Illinois University School of Medicine in Springfield, who has led several of the past phase II trials testing IA delivery, said that although he gives kudos to the Dutch team for taking on the study, the exact methods they used for the trial might have limited the efficacy of IA delivery.
Specifically, when patients had a tumor confined to one side, the physicians delivered the whole chemotherapy dose through the carotid artery on that side. However, Dr. Rasch’s team decided that if a patient’s tumor crossed the midline, they would split the chemotherapy dose in half, delivering 50% through each carotid artery.
Theoretically that could have led to undertreatment of the bulk of the disease, said Dr. Robbins, who discussed Dr. Rasch’s study during the ASTRO plenary session. First, most of these tumors would not be distributed equally on the two sides of the midline. Second, even for tumors that have a significant portion of their bulk across the midline from their point of origin, most of their blood supply will continue to come from the artery on the side where the disease originated.
A post-hoc subset analysis of patients in the trial may support that theory. Our analysis is not complete, but early results suggest that patients with a lateralized tumor and given unilateral infusions had better results, Dr. Robbins said. The interpretation of that observation is complicated, though, by the fact that unilateral tumors also tend to be smaller in volume than bilateral ones. And because tumor volume is known to affect the success of treatment in head and neck cancer, it is not immediately apparent whether the patients benefited from unilateral infusion or from smaller tumor size at the time of treatment.
In his own clinical experience, Dr. Robbins said that he treats 90% of his patients with a unilateral infusion, even if their disease has crossed the midline. Only when patients have a tumor spread nearly equally between the two sides would he use bilateral infusion. By contrast, approximately half of the patients in the Dutch trial were treated with bilateral infusion.
A number of us still believe this is a good way to treat patients, Dr. Robbins said. With that in mind, he and colleagues in the United States are currently designing a randomized controlled trial to test IA administration again. This time, however, the protocol will specify that unilateral infusion is to be used in most cases. The team will also look at tissue samples before and after treatment to determine how much drug is getting into the tissue and what effect it is having on the tumor.
In theory, IA delivery should lead to higher levels of drug in the tumor relative to IV doses. Because the drug is introduced in close proximity to the tumor, the diseased tissue is exposed to higher concentration of the agent, relative to what it would be exposed to with systemic IV delivery. Additionally, by pairing IA drug delivery with IV administration of sodium thiosulfate, which scavenges excess drug from the blood, physicians can give patients higher doses of the drug without risk of extensive side effects, as would be seen if such doses were delivered IV.
In fact, a reduction in side effects was the one area in which Dr. Rasch’s group found significant differences between the two trial arms. Only 1% of patients in the IA arm had renal toxicity as a result of therapy, compared with 10% in the IV arm. Also, patients in the IA arm had less trouble with nausea and vomiting. On the other hand, 9% of IA patients had neurological toxicity, compared with just 1% in the IV arm.
For Dr. Robbins, other benefits far outweigh the risk of neurological toxicity-particularly because most patients recover from neurological complications in both his experience and in the Dutch trial.
©2007 The Triological Society