Although glucocorticoids are often used in cancer therapy, in particular to enhance the effectiveness of antiemetic therapy, they have been associated with impaired tumor apoptosis and an increased frequency of metastases in some reports. The current study aimed to determine whether glucocorticoid treatment had an adverse effect on outcomes in patients with ovarian carcinoma.
Records of patients with ovarian carcinoma who were scheduled to receive at least six courses of systemic chemotherapy were reviewed. Patients were grouped into those who had or had not received corticosteroid medication as a part of general antiemetic prophylaxis before chemotherapy, and details of hematologic parameters during treatment and disease recurrence-free and overall survival were recorded.
Altogether, 245 patients with ovarian carcinoma had received chemotherapy. Of these, 62 had been given concurrent glucocorticoid treatment and 183 had not. The two patient groups were well balanced with respect to disease stage and other prognostic factors. Kaplan–Meier analyses showed no significant differences in survival between the groups. Patients who received glucocorticoid treatment had significantly higher leukocyte values in the days immediately after chemotherapy, higher nadir leukocyte values, and higher counts before subsequent courses of chemotherapy (P < 0.01; Levene test, t test) compared with patients who did not receive glucocorticoid treatment. As a result, the initial treatment targets were achieved significantly more often in the glucocorticoid group (P = 0.007; chi-square test).
Glucocorticoids are used widely in medicine, ranging from the treatment of inflammatory disorders (e.g., rheumatoid arthritis, asthma, dermatitis) and autoimmune processes (e.g., Crohn disease, idiopathic thrombocytopenia) to prophylactic treatment to achieve better maturation of fetal lungs in preterm neonates. In oncology, glucocorticoid treatment has also proved beneficial in several situations, including the following: 1) reduction of operative morbidity1; 2) improved effectiveness of antiemetic drugs (e.g., benzamides, 5-HT3 antagonists)2–5; 3) substitution in cases of treatment with the early aromatase inhibitor aminogluthimid6; 4) reduction of tissue edema in cases of brain metastases and cranial radiotherapy, as well as prevention of radiation-induced emesis7–9; 5) treatment of leukemia, lymphomas, and hormone-resistant advanced prostate carcinoma10–12; 6) prevention of the side effects of cancer drugs, for example, allergic reactions related to taxanes and pulmonary toxicity syndrome with 1,3-bis (2-chlorothyl)-1-nitosourea13, 14; and 7) improvement in the quality of life of patients with terminal cancer.15, 16
Concerns about the widespread use of glucocorticoids during therapy for solid tumors have been expressed repeatedly for more than 45 years.17–20 Some published reports have questioned the safety of glucocorticoids in patients with cancer. However, these concerns do not allude solely to the known side effects of these drugs—electrolyte disturbances, cardiovascular effects, diabetes mellitus, and loss of bone density and osteoporosis.21–23 In 1958, Iversen and Hjort17 published the results of an autopsy study on patients with disseminated breast carcinoma. In that study, 26% (8 of 31) of women who received glucocorticoid treatment had spleen metastases, whereas none of those (0 of 16) who had not received the therapy developed spleen metastases. Further studies showed significant increases in metastatic spread to the lung, liver, heart, opposite breast, brain, spleen, and the submucosa of the gastroduodenum.18
These data also have been supported by experimental work.24 In patients with bronchial carcinoma, cortisone treatment was associated with a poor prognosis.25 More recent data suggest a negative influence of endogenous cortisol concentrations on the incidence and prognosis of patients with renal cell carcinoma.26 Further in vitro research found reduced natural cytotoxic cell activity and inhibited apoptosis or induction of apoptosis resistance in cancer cells under the influence of glucocorticoids.20, 27, 28 In addition, glucocorticoid receptors are suspected of having a role in carcinogenesis by activating H-ras oncogenic potency.29
Finally, further arguments against the use of glucocorticoids are provided by researchers in the field of psychoneuroimmunology. Cortisol is one of the most important stress hormones. It is believed that use of glucocorticoids in patients with high stress may enhance immunosuppression and increase comorbidity and, perhaps, negatively influence survival in these patients.30–32 Certainly, these effects are not well established and may be the result of the interaction of personality variables, such as anxiety, and long-term glucocorticoid treatment.
Based on the arguments against the use of glucocorticoids, skepticism regarding the use of dexamethasone and other glucocorticoids seems justified. Unfortunately, no prospective clinical studies have yet addressed this topic. We undertook this retrospective study to elucidate the effects of corticosteroids on hematologic parameters during chemotherapy, on the response to chemotherapy, and on survival in patients with ovarian carcinoma.
MATERIALS AND METHODS
The current study evaluated the charts of patients with histologically confirmed ovarian carcinoma who were treated between 1985 and 1998 at the Department of Obstetrics and Gynecology, Justus-Liebig-University (Giessen, Germany). Only patients who were scheduled to receive at least 6 courses of systemic chemotherapy (cisplatinum 50 mg/m2, epirubicin 60 mg/m2, and cyclophosphamide 500 mg/m2; 4 courses over 28 days) were included. Patients who received other regimens or no chemotherapy were excluded, as were patients who had concomitant neoplasms.
The clinical prognostic factors assessed included the following: the patient's age, tumor stage, residual tumor size, completion of chemotherapy, and hematologic parameters during treatment. We did not reevaluate tumor histology and grading because the initial surgical specimens were unavailable for many patients.
Most patients were seen in the hospital outpatient department every 3 months within the first 3 years after diagnosis, every 6 months between 3–5 years after diagnosis, and annually thereafter. Study patients lost to follow-up were checked out routinely every 6–12 months through the government data systems to ascertain their survival status. Survival status was reevaluated in December 2003. Outpatient follow-up included patient's history, physical examination, and, from 1987 onward, the determination of the tumor marker CA 125.33
Data Management and Statistics
The primary end points of the study were hematologic toxicity, overall survival, and disease recurrence-free survival, which were calculated with respect to the effects of various clinical parameters using the t test of independent samples (including the Levene test) and Kaplan–Meier analysis. Data management and statistical analyses (bivariate correlation and Kaplan–Meier survival analyses, residual analysis) were performed using SPSS software for Windows 11.0 (SPSS, Inc., Chicago, IL). A probability of error < 5% was regarded as significant.
Of the 763 patients with ovarian carcinoma surveyed, 245 qualified for inclusion in the current study. Patient characteristics (age, Karnofsky index, stage according to the International Federation of Gynecology and Obstetrics), histology and grading, residual tumor volume, treatment plans, dose intensity, follow-up time, and survival status) are shown in Table 1. The 245 women were divided into 2 groups—those who had received glucocorticoids before chemotherapy to potentiate the antiemetic efficacy of benzamides or 5-HT3 antagonists (n = 62) and those who had not received glucocorticoid treatment (n = 183). Patients who received continuous glucocorticoid treatment for other reasons were excluded from the study. No statistically significant differences were found between the groups (using the chi-square test). Most patients who received glucocorticoids were given 8 mg or 20 mg of dexamethasone (51% and 32%, respectively) or comparable doses of other glucocorticoids (17%).
Table 1. Patient and Tumor Characteristics in the Two Treatment Groups
No glucocorticoid treatment (n = 183) (%)
Glucocorticoid treatment (n = 62) (%)
FIGO: International Federation of Gynecology and Obstetrics.
Univariate Kaplan–Meier analyses of overall and disease recurrence-free survival showed no statistically significant differences between the groups (Figs. 1 and 2). However, glucocorticoid treatment versus no glucocorticoid treatment was associated with better overall clinical response rates and higher rates for completing the treatment schedule as planned (Table 2).
Table 2. Outcome of Treatment in the Two Patient Groups
No glucocorticoid treatment (n = 183) (%)
Glucocorticoid treatment (n = 62) (%)
χ2 (df; P)
df: degrees of freedom.
Treatment competed as planned
7.2 (1; 0.007)
Overall response to treatment
8.6 (3; 0.035)
Before chemotherapy, the distributions of leukocyte and thrombocyte levels were similar in the two groups. However, significant differences between the groups were observed for each course of chemotherapy as well as subsequent courses. In general, patients receiving glucocorticoids had higher leukocyte counts on Days 2 and 3, higher nadir values, and higher levels before the start of the next treatment course (Table 3). With respect to thrombocyte values, we observed trends favoring the administration of glucocorticoids, but these were not as pronounced as with leukocyte values (P < 0.10). A significant protective effect was observed in the hemoglobin concentrations in patients given glucocorticoid prophylaxis (Table 4).
Table 3. Course of Leukocyte Counts before and on Day 2 or 3 after Chemotherapy and Nadir Leukocyte Counts in Relation to Administration of Glucocorticoids for Antiemetic Prophylaxis
Did not receive glucocorticoids
No. of patients
Mean leukocyte count (cells/μl)
No. of patients
Mean leukocyte count (cells/μl)
Difference between groups was statistically significant (t test: P < 0.01).
Our study failed to show any adverse effects of glucocorticoids in patients with ovarian carcinoma. There was no evidence whatsoever that glucocorticoids had a detrimental effect on the efficacy of antiemetic drugs, the overall response to treatment, or patient survival. On the contrary, our analyses showed that giving corticosteroids to improve the effectiveness of antiemetic drugs may actually protect the bone marrow so that patients receiving dexamethasone have better leukocyte counts throughout each course and until the next course of chemotherapy, enabling them to adhere to the treatment schedule. Glucocorticoids seem to protect megakaryocytes or thrombopoieis to a lesser extent, whereas the reduction in hemoglobin concentrations during courses of chemotherapy may have been slowed. Unfortunately, we were unable to analyze any other parameters of toxicity.
With regard to the protective effect of corticosteroids on the hematopoietic system, our findings confirm results from preclinical studies as well as a very small randomized trial.34–39 The small randomized trial studied pretreatment with dexamethasone, versus granulocyte-macrophage–colony-stimulating factor treatment, versus no pretreatment. However, the study cohort comprised only 28 patients who received many chemotherapy regimens. These patients had a variety of tumors, including sarcomas, and received various pretreatment protocols.39 Data on the course of leukocyte values after glucocorticoid treatment are well in line with findings from obstetrics studies, where the number of circulating leukocytes and granuloocytes increased after administration of glucocorticoids (e.g., betamethasone).40–42
In contrast to the studies against dexamethasone therapy, others have reported the potential benefits of dexamethasone therapy. In vitro studies found that dexamethasone inhibited ovarian carcinoma cell proliferation and induced differentiation in a time-dependent and dose-dependent manner.43–45 There are also reports on benefits during palliative treatment of prostate carcinoma.46 However, the mechanisms of cellular glucocorticoid action are not fully understood. In prostate carcinoma, glucocorticoids seem to interfere with the NF-κB–interleukin-6 pathway via nuclear glucocorticoid receptors, whereas in lymphomas and leukemia, a membrane-associated glucocorticoid receptor was found to be associated with cell lysis.47, 48 Immunohistochemical studies in colon carcinoma found positive correlations with the cell cycle-related molecules pRb and p16.49 Although it did not reach statistical significance, survival in patients with glucocorticoid receptor-positive tumors was somewhat better than survival in patients with other tumors.49
This retrospective analysis cannot give a definitive answer to the question of whether glucocorticoid treatment, given as a part of antiemetic prophylaxis or prevention of allergic reactions to taxanes, is safe in patients with ovarian carcinoma. However, until further studies have addressed this topic, our results can help to assuage fears.
We believe that this topic deserves more attention to remove or justify the skepticism surrounding glucocorticoid treatment, particularly in relation to continuous treatment for other reasons than antiemesis. However, these studies may face difficulties because of the established benefits of glucocorticoids and should, of course, include other types of tumors.