Jean-Pierre, P., Morrow, G. R., Roscoe, J. A., Heckler, C., Mohile, S., Janelsins, M., . . . Hopkins, J. O. (2010). A phase 3 randomized, placebo-controlled, double-blind, clinical trial of the effect of modafinil on cancer-related fatigue among 631 patients receiving chemotherapy: a University of Rochester Cancer Center Community Clinical Oncology Program Research base study. Cancer, 116, 3513–3520.

To examine the effect of modafinil on patient-reported fatigue in patients with cancer who were undergoing chemotherapy.

Assessments were conducted at baseline after randomization and shortly after cycle two of therapy. Modafinil or placebo was started at 100 mg on day 10 or day five of study cycle two, then increased to a full dose of 200 mg after three days. This regimen then was continued until day seven of treatment cycle four, at which time all patients discontinued medication.

• A total of 877 participants were enrolled, and 631 were analyzed.
• Mean ages for the study groups were 60 and 61 years (range 18–90), and 66% to 68% were female.
• The most common sites were gastrointestinal, breast, and lung. The sample also included genitourinary, gynecologic, hematologic, and other cancers.
• Participants were required to be beginning a cancer treatment course of at least four planned cycles of chemotherapy with at least two weeks apart, with no concurrent radiation or interferon treatment.
• Patients were excluded if they had taken modafinil or any psychostimulant within the past 30 days.
• Participants had at least a score of 2 on the Brief Fatigue Inventory (BFI) question 3 (worst level of fatigue).
• Of the patients, 57% had received prior chemotherapy, and 22% to 24% had received prior radiation therapy.
• Most (67%–70%) participants were married, and about half had some college education.

This multisite study was set in 23 geographical areas across the United States among University of Rochester Cancer Center Community Clinical Oncology Program (URCC CCOP) affiliates.

This was a randomized, placebo-controlled, double-blind trial.

• BFI, question 3
• Epworth Sleepiness Scale (ESS) to measure excessive daytime sleepiness
• Profile of Mood States depression subscale (POMS-DD)
• Missing scores at cycle four were replaced with scores from cycle three for those who completed the study and only had evaluable data through cycle three (n = 58 for modafinil; n = 29 for placebo).

ANCOVA for BFI fatigue score showed an interaction between treatment effects and baseline BFI score (p = 0.017). A significant difference existed between the study groups for those who had severe fatigue at baseline (BFI of 7 or greater), with average score in the modafinil group. No differences in fatigue were observed between the study groups for those who had mild or moderate baseline fatigue. Daytime sleepiness on ESS showed significant improvement in the modafinil group (p = 0.002). No significant differences existed in depression outcomes between the groups. In the modafinil group, 11% of patients experienced adverse events, and in the placebo group, 9% had adverse events. Only three adverse events were judged to be definitely associated with treatment with modafinil: allergic reaction, dyspnea, and headaches.

The findings supported the use of 200 mg of modafinil as an effective treatment for severe cancer-related fatigue in patients undergoing chemotherapy. Modfinil was not effective for patients with less severe fatigue.

• It is not clear what the statistical effect was of the replacement of missing cycle four data with data from cycle three. The authors analyzed multiple data replacements to test this and found no differences. This suggests that the time frames of the study measures were irrelevant.
• The study demonstrated effectiveness for the short term during active treatment with only chemotherapy; the findings may not be the same for other groups of patients or other time frames of observation.
• Diversity of the sample in terms of racial and some other demographic findings was limited.
• Although no significant differences existed between study groups that would have affected the study results, the findings may not be applicable to other patients with different demographic characteristics.