Randomized trials of induction chemotherapy. A critical review

There are nine prospective randomized trials comparing induction chemotherapy with standard therapy published in either final form or as preliminary abstract reports. Only one of the six randomized trials critiqued was designed to account for all known prognostic factors, utilized an effective chemotherapy regimen, and had excellent compliance such that the numbers of patients completing the protocol were adequate to provide a valid statistical interpretation of the data. This was the VA cooperative group trial to preserve the larynx. Although this trial did not show a difference in survival between surgically treated patients and those who received induction chemotherapy, the larynx was preserved in two thirds of patients. The results of the other published trials cannot be considered conclusive owing to the flaws in design and interpretation noted in this review. These trials do confirm the feasibility of administering chemotherapy prior to surgery and RT and do confirm the prognostic importance of various factors suggested by the results of single institution trials. Induction chemotherapy has been tested for almost two decades. During this time we have learned the importance of treating intensively to obtain a high complete response rate. At one primary site, the larynx, it has been shown that 64% of patients can be rendered histologically disease-free after three courses of cisplatin and 5-FU. The results of three randomized trials indicate that induction chemotherapy can change the expected pattern of recurrence by decreasing the rate of distant metastases. This early systemic treatment of micrometastases may also reduce the mutation rate theorized for the development of drug resistance. The goals of future randomized trials should include the use of dose-intensive multidrug regimens to increase complete response rates, improve locoregional control, and decrease distant metastases. Induction chemotherapy trials with cisplatin and 5-FU should explore organ preservation at sites other than the larynx where significant functional impairment results from standard surgical approaches. Randomized trials must be carefully designed and rigorously conducted to ensure that the results and conclusions are valid. New regimens, perhaps incorporating growth factors, need to be identified and tested in this group of patients. Finally, improvement in survival must remain a primary goal of multimodality therapy for advanced head and neck cancer.