Comparison of outcomes in studies of advanced Hodgkin's lymphoma

Johnson, Peter W.M., Radford, John A., Cullen, Michael H., Sydes, Matthew R., Stenning, Sally P. and Hancock, Barry W. (2006) Comparison of outcomes in studies of advanced Hodgkin's lymphoma. Journal of Clinical Oncology, 24 (20), pp.3309. (doi:10.1200/JCO.2006.06.2976).

Abstract

We were interested to read the editorial by Dr Carde1, which referred to our article2 that reported the results of the United Kingdom Lymphoma Group LY09 study in advanced Hodgkin's lymphoma. The editorial may have given the impression that our study was a randomized comparison of three regimens, which is not the case. A standard arm of doxorubicin, bleomycin, vinblastine, and dacarbazine (ABVD) was used, and investigators elected in advance whether to randomize against the alternating or hybrid regimen. Thus, it is not legitimate to make direct comparison of the two multidrug arms. It is also difficult to compare the overall results in this study to the Italian trial reported in the same issue by Gobbi et al3, which included a more favorable group of patients among whom only 11% of patients were in International Prognostic Score group 4 to 7, compared with 19% in the UK study. The measure of 65% progression-free survival at 5 years cited by Dr Carde refers only to patients in stages III/IV, whereas for the UK study as a whole the figure is 73%.
The interpretation of information regarding the use of radiotherapy in the two parallel randomizations is problematic. Dr Carde suggests that a comparison can be made between the two standard ABVD groups, with the more frequent use of eight cycles in the centers randomizing against alternating treatment apparently giving marginally better freedom from progression results than those seen in the hybrid randomization, where it was more common to use six cycles followed by radiotherapy. Such data must be interpreted with caution, especially because the patients in the hybrid randomization were more likely to have bulky mediastinal disease and systemic symptoms, suggesting a worse prognostic group. We contend that such an effect may be due to patient selection rather than the use of radiotherapy or fewer cycles of ABVD.
Finally, Dr Carde has made a provocative analysis of the data, correlating freedom from progression with intended, rather than actual doses delivered. This may be helpful in terms of generating hypotheses to be tested in future studies but does not necessarily explain observed outcomes in the LY09 trial.

This record has no associated files available for download.

Contributors

Author: Peter W.M. Johnson

Download statistics