W e read with great interest the recent article by Shin et al. 1 Dr. Shin and his team have performed a very interesting Phase II study; however, our ability to assess the results of the treatment are limited by data missing in the report. It would be useful and informative to know how many patients experienced neutropenia, how long neutropenia lasted, and what grade of neutropenia was encountered. It would also help to know how many patients experienced a delay in chemotherapy or required dose reductions because of neutropenia and/or thrombocytopenia, since these parameters are needed in assessing the toxicity of a treatment. While we are given the median time of followup, the range of follow-up is missing. These data are important in interpreting the survival data provided. In addition to these basic questions, there are other issues of care and treatment which impact efficacy and may reflect treatment-related toxicity. For example, it would be important to have an analysis of important parameters such as the median time to starting radiation after induction and after surgery when surgery was used first. We believe these are important parameters in a chemotherapy driven organ preservation approach.We also have some concerns about using surgery between chemotherapy and radiation when chemotherapy is being used for organ preservation. This has the potential of delaying radiation therapy and permitting tumor repopulation in the surgical bed. 2-4 The rationale behind using clinical partial response (PR) at the primary site as a surrogate for delaying radiation therapy in these patients is unclear and the findings at surgery in these patients are not presented. The assessment for response after induction chemotherapy in that article was not a pathologic assessment based on primary site biopsy. In our experience visual inspection in partially responding patients can be misleading, and we routinely perform a biopsy prior to starting radiation. This might be of more prognostic value and was documented in a Veterans Affairs study. 5 Also, while of minor importance, the reported survival of 83% does not include the two patients who died after being entered on the study and treated. With an intent to treat analysis these treatment related deaths should be included in the overall survival figure, which decreases to 79%.Finally, the value of ifosfamide in head and neck carcinoma also is unclear, and whether this drug adds to the response rate achieved with carboplatin and paclitaxel alone is debatable. The responses reported to different carboplatinum/taxol regimens in locally advanced, previously untreated disease are quite high. 6