Few things in oncology are as complex—and sensitive—as corticosteroid-induced (hence, iatrogenic) resistance to treatment of solid tumors. Inducible mechanisms that confer resistance may be cellular, biochemical, or molecular, including the induction of enhanced DNA damage repair as well as apoptosis inhibition.1, 2 Moreover, corticosteroid-induced diabetes may cause biophysical resistance while promoting the metastatic process.2, 3

Induction of resistance and promotion of metastasis by concomitant administration of a corticosteroid during chemotherapy were reported first almost 5 (!) decades ago, yet only recently has a third clinical study followed.4 Its authors clearly deserve broad applause for their efforts. Strikingly, they discovered that corticosteroids also protect the bone marrow.

Their other finding—no evidence for the induction of treatment resistance—should be viewed with some skeptical distance, however. First, theirs was not a prospective, randomized trial, unlike the recently published report on a trial that addressed corticosteroid use in the management of brain injury.5Second, the study evaluates only 245 of 763 patients with ovarian carcinoma surveyed; thus, one may wonder about the information in the 518 omitted charts. Third, the findings shown in their Figure 2 of obvious but statistically nonsignificant differences regarding tumor control well may result from the small sample size rather than from the absence of a true difference. Forth, the absence of an impact on overall survival does not preclude at all an impact of corticosteroids on the tumors. Fifth, the fact that corticosteroids protect the bone marrow is noteworthy; however, is not much of an advantage in this age of molecular oncology, in which specific agents to protect and rescue bone marrow are available readily.

In conclusion, almost half a century after initial reports on corticosteroid-induced treatment resistance and metastasis promotion, we still are dearly in need of an in-depth analysis. Obviously, addressing these difficult issues by Münstedt et al.4 and by editors demonstrates courageous dedication to dealing with these problems.

However, they should not be left alone. There are virtually thousands of patient files waiting to be (re-) evaluated, and the problems regarding the planning of an eventual prospective clinical trial, put simply, are manifold.1 Nevertheless, hundreds of thousands of patients undergo chemotherapy every year, and most of them are treated concomitantly with a corticosteroid, primarily with the objective of reducing chemotherapy-associated nausea and hyperemesis. Given these numbers, even small relative differences in tumor response may translate into huge absolute effects; hence, the issue merits appropriate analysis at last.


  1. Top of page

Hans Peter Rutz M.D.*, Ingrid Herr Ph.D.†, * Division of Radiation Medicine, Proton Therapy Program, Paul Scherrer Institute, Villigen, Switzerland, † Clinical Cooperation Unit, Molecular Oncology, German Cancer Research Center, Heidelberg, Germany.