The reservations expressed by Dr. Herr and Dr. Rutz regarding our report, “Glucocorticoid administration in antiemetic therapy: is it safe?”1 do not seem applicable, because the value of retrospective studies generally is underestimated, and the sample size is sufficient to exclude clinically relevant effects. Furthermore, the bone marrow-protective effect of glucocorticosteroids may be of some importance, because recombinant growth factors are not available throughout the world due to cost issues, because there also has been concern about the tumor-protective effects of growth factors, because not all patients respond to growth factors, and because, to date, no factors for the stimulation of thrombopoiesis are available. Finally, the referenced article may not provide a definitive answer to all open questions regarding the issue of glucocorticoid treatment in patients with cancer, but it does contribute in general to the acquisition of knowledge on the topic and may prevent some physicians from becoming too confused by mere in vitro data.
First, we understand the desire for a prospective, randomized study. However, the value of carefully conducted, retrospective studies generally is underestimated.2, 3 Second, not all patients could be included, because patients with disease in very early stages simply did not require chemotherapy. Other patients received different therapeutic regimens or alternations between radiotherapy and chemotherapy, other patients received primary palliative strategies, whereas some patients refused treatment. However, prospective, randomized trials also do not include all patients; in fact, all criteria for inclusion and exclusion represent factors that create potential biases.
Third, 214 patients cannot be considered a small number, especially in view of the background of large numbers of events and careful patient selection. We do not deny an effect of glucocorticosteroid on tumor cells; however, the in vitro phenomena reported to date may not reflect the in vivo situation or may not have clinical importance. There also may be important differences between long-term glucocorticosteroid administration and single-shot treatment as antiemetic therapy.
Fourth, we agree that, in the past few years, recombinant granulocyte-colony stimulating factor (G-CSF), granulocyte–macrophage-colony stimulating factor (GM-CSF), and erythropoietin have lead to greater patient safety and risk reduction after chemotherapy. However, four major aspects have not been considered by Dr. Herr and Dr. Rutz: 1) Recombinant G-CSF, GM-CSF, and erythropoietin are not available readily in all parts of the world; 2) there have been discussions about the safety of growth factor administration4, 5; 3) erythropoietin may be available, but numerous patients will still require blood transfusions because of response rates; and 4) chemotherapy also affects thrombopoiesis. No recombinant growth factor has yet been proven both safe and efficient.
In conclusion we agree with Dr. Herr and Dr. Rutz that, before our study, no measures had been taken to examine this problem diligently in the clinical situation.6 We are aware of the limitations of our study, as discussed in the report, and of the fact that the study provided no definitive answers to the open questions regarding the issue of glucocorticoid treatment in cancer patients. However, our work shows that the findings obtained by in vitro studies do not necessarily translate into the situation in vivo and that that glucocorticosteroids not only may improve both tolerance of chemotherapy and quality of life, but they also may improve outcomes. We believe that further clinical studies and retrospective analyses on this issue are needed badly, but clinicians also should not be confused by mere in vitro data.