The effect of modafinil on cognitive function in breast cancer survivors†‡
This study was presented as part of the Patient and Survivor Care Oral Session at the American Society of Clinical Oncology Annual Meeting, Chicago, Illinois, June 1-5, 2007.
The authors are indebted to all the volunteers for participation and to Cephalon for providing the study drug.
Abstract
BACKGROUND:
The authors conducted a randomized clinical trial examining the effects of modafinil in reducing persistent fatigue in patients after treatment for cancer and performed secondary analyses to assess the effect of modafinil on cognitive function.
METHODS:
Breast cancer patients who reported a score of ≥2 on the Brief Fatigue Inventory were enrolled in the study. In phase 1 (P1), patients received 200 mg modafinil open-label once daily for 4 weeks. In phase 2 (P2), patients with a positive response after P1 were randomized either to an additional 4 weeks of modafinil or to placebo. Tests of memory and attention selected from the Cognitive Drug Research (CDR) computerized cognitive assessment were performed at baseline (before modafinil) and after completing phases 1 and 2. The paired differences for each test score were subjected to a Wilcoxon signed rank test.
RESULTS:
Of the 82 women who were enrolled, 76 completed P1, and 68 completed all assessments in the study. Modafinil had a significant effect on the Speed of Memory (P = .0073) and Quality of Episodic Memory (P < .0001) during P1 of the study. After randomization at Week 8, those patients who continued modafinil demonstrated significantly greater improvement in Speed of Memory (P = .029), Quality of Episodic Memory (P = .0151), and mean Continuity of Attention (P = .0101) relative to the group that was switched to placebo.
CONCLUSIONS:
The authors found that modafinil improved cognitive performance in breast cancer survivors by enhancing some memory and attention skills. Although confirmation is needed, these findings suggest that modafinil may enhance quality of life in this patient population. Cancer 2009. © 2009 American Cancer Society.
After cancer treatment, some breast cancer survivors report cognitive impairment, which has commonly been attributed to the receipt of chemotherapy and colloquially termed “chemobrain.”1 Cognitive deficits identified by cancer survivors consist of a range of difficulties including memory and concentration problems that can emerge during cancer treatment and continue to have long-term effects for many months and years after its completion.2-8 Impairment of brain function profoundly affects cognition, psychological well-being, and the ability to perform the usual activities of daily living.
Changes in cognitive function may be confused with or confounded by other problems commonly associated with cancer and its treatment, such as depression, anxiety, and fatigue caused by sleep deprivation.9-11 Although the phenomenon is not understood completely, cognitive deficits in cancer patients are assuming greater significance as cancer survival improves. Advances in basic sciences, imaging, and clinical sciences are beginning to unravel pathophysiologic mechanisms that may underlie this problem.12 Pharmacologic treatment options for cognitive deficits in cancer survivors are borrowed from diverse diseases, including attention-deficit/hyperactivity disorder and neurodegenerative diseases.13, 14 Conventional therapies soon may find new applications; for example, recent preclinical data suggest that modafinil (Provigil) may have some cognitive-enhancing abilities, which may positively affect patients who experience “chemobrain.”
Modafinil is a novel wake-promoting agent that is effective and well tolerated in the treatment of excessive sleepiness associated with narcolepsy and in persons with shift-work sleep disorder.15-19 The precise mechanism(s) through which modafinil promotes wakefulness is unknown. Modafinil has wake-promoting actions like those of sympathomimetic agents, including amphetamine and methylphenidate, although the pharmacologic profile of modafinil is not identical to that of the sympathomimetic amines. The recommended single daily dose range for narcolepsy is 100-400 mg modafinil. These doses have wake-promoting efficacy comparable to 20-40-mg doses of methylphenidate,20 which has been used previously in cognitive studies in patients with advanced cancer.21
We completed a study examining the effect of modafinil on fatigue and sleep complaints in 82 breast cancer patients after completion of chemotherapy.22 Patients received 200 mg modafinil once daily for 1 month. Of the 82 women enrolled, 76 completed the study and began the study medication. A majority, 68 of the 76 women (90%) who completed the 1-month study, reported improvement on all 4 measures of fatigue used,23-26 all P < .001 (Fig. 1). In addition, the Epworth Sleepiness Scale showed a 22% decrease (P < .01), and the “disturbed sleep” symptom on the symptom inventory27 decreased >45% (P < .01).
Fatigue levels in posttreatment breast cancer patients receiving open-label modafinil.22
Despite these findings, little is known about the causes of cognitive changes seen during and after cancer treatment. It seems reasonable to assume that they are related, at least in part, to other psychological and biological treatment complaints such as fatigue, sleeping problems, and depression.10, 28 The nature of the relationships among these variables, however, both at onset and over time, remains unclear.
To our knowledge, despite evidence that modafinil acts as a cognitive enhancer in healthy subjects,29-31 no study to date has examined the effect of modafinil on a wide range of cognitive functions using a comprehensive and well-validated neuropsychological test battery in cancer patients. Therefore, as modafinil was effective in reducing persistent fatigue in our primary study,22 we conducted secondary analyses to assess the effect of modafinil on cognitive function and to examine whether modafinil might act as a cognitive enhancer in breast cancer patients who have completed treatment for cancer.
MATERIALS AND METHODS
General Study Procedures
Eighty-two breast cancer patients between the ages of 33 and 83 years, who had had chemotherapy and/or radiation treatment at least 1 month previously for an initial diagnosis of cancer, were able to swallow medication, had a score of 2 or greater on a Brief Fatigue Inventory (BFI)25 question of “fatigue worst,” and had completed participation in the primary study “Modafinil to Treat Persistent Fatigue in Patients Following Treatment for Cancer”22 are included in this secondary analysis. These patients were approached for the primary study by their medical oncologist either at a regularly scheduled appointment at the Medical Oncology clinic at the University of Rochester in Rochester, New York, by telephone, or by a standardized letter in the mail. It was explained to the patients that fatigue is a subjective symptom with individual meaning that may be described using a variety of words and phrases such as “weak and tired,” “no energy,” “can't get motivated,” or “trouble breathing.”32 The patient then completed question 3 of the BFI. If the patient responded to question 3 of the BFI with an answer of ≥2, the patient was invited to participate in the study; however, if the patient responded with an answer of 1, the patient was considered ineligible for the study. The University of Rochester Research Subjects Review Board approved the primary study, and all patients provided written informed consent.
In phase 1 of the study, patients received 1 tablet of 100 mg of modafinil once a day for the first 3 days and then 2 tablets (200 mg) open-label once daily for 4 weeks. In phase 2, patients with a positive response after phase 1 were randomized either to an additional 4 weeks of modafinil (200 mg) or to placebo. Tests of memory and attention selected from the Cognitive Drug Research (CDR) computerized cognitive assessment were performed at baseline (before modafinil) and after completing phases 1 and 2.
Potential research subjects were excluded if they had taken modafinil previously, had taken a psychostimulant or a monoamine oxidase inhibitor within the past 30 days, or were currently taking an anticoagulant, α-interferon, interleukin-2, or a corticosteroid. Individuals with a history of clinically significant cardiac disease, uncontrolled hypertension, alcohol or drug abuse, severe headaches, glaucoma, seizure disorder, narcolepsy, psychotic disorder, or uncontrolled Tourette syndrome were also excluded. If a steroidal contraceptive was being used for fertility control, participants had to agree to use a barrier method of contraception during the study and for 1 full menstrual cycle after the study.
Randomization
After 4 weeks of open-label modafinil, patients who met the eligibility criteria were randomized to 1 of 2 trial arms by means of randomly ordered sealed envelopes. Participants were stratified by treatment regimen (completed chemotherapy alone or radiation alone or chemotherapy plus radiation) and were assigned either to continue 200 mg modafinil orally once each day or to switch to a matching placebo orally once each day.
Assessments
The CDR system was used to assess major domains of attention and memory. The CDR system is designed to provide the drug development process with a stable, practical, validated, and sensitive platform to enable important cognitive and behavioral effects of medicines to be captured and evaluated. Its validity, test-retest reliability, and sensitivity have been demonstrated in several populations.33-36
The CDR system comprises both custom-built tests (eg, digit vigilance, spatial working memory) and tests whose general methodology is widely known (simple and choice reaction time, word recall). All tests have been standardized, extensively validated, and implemented with a software and hardware platform that enables responses to be recorded with true millisecond accuracy.
The system has been used worldwide since 1984 in over 500 clinical trials. Data from the system have been the subject of over 100 peer-reviewed papers, more than 20 book chapters, and 300 published abstracts.37 An advantage of the CDR system is that many of the tests used are generic and generally accepted for use in psychopharmacology research. Furthermore, because it has been used in hundreds of studies, extensive normative data exists from populations ranging in age from 12 to 90 years.
In recent years, CDR has identified major factors reflecting distinct and important dimensions of human cognitive function. The measures that contribute to these factors are combined to produce factor scores. These factor scores are unique to the CDR system, and thus the CDR offers a highly sensitive way of assessing broad but distinct aspects of cognitive function.
The sensitivity and reliability of these tests for detecting cognitive impairment in a variety of clinical conditions (eg, normal aging, depression, poststroke, schizophrenia, and attention deficit disorder, all of the major dementias, head trauma, and drug-induced cognitive changes) have been demonstrated.38-45 The CDR battery is sensitive to treatment effects, including pharmacologic therapies.34, 36, 44-51
Table 1 shows a summary of the measures contributing to the analyses of the CDR computerized assessment system.
| Factor Score and Mean Normative Data for Women Aged 33-83 y | Measures That Contribute to the Score |
|---|---|
| Quality of Episodic Secondary Memory (lower score is better), mean: 199.3 | Immediate word recall—word correctly recalled |
| Immediate word recall—intrusions | |
| Immediate word recall—errors | |
| Delayed word recall—word correctly recalled | |
| Delayed word recall—intrusions | |
| Delayed word recall—errors | |
| Word recognition, new stimuli—accuracy | |
| Word recognition, original stimuli—accuracy | |
| Picture recognition, new stimuli—accuracy | |
| Picture recognition, original stimuli—accuracy | |
| Quality of Working Memory (lower score is better), mean: 1.557 | Numeric working memory |
| Speed of Memory (lower score is better), mean: 3089 | Speed of numeric working memory |
| Speed of word recognition | |
| Speed of spatial working memory | |
| Speed of picture recognition | |
| Power of Attention (higher score is better), mean: 1201 | Simple reaction time |
| Choice reaction time | |
| Digit vigilance—speed | |
| Continuity of Attention (higher score is better), mean: 89.71 | Choice reaction time—accuracy |
| Digit vigilance—targets detected | |
| Digit vigilance—false alarms |
Statistical Analysis
The main objective of this study was to evaluate the efficacy of modafinil in alleviating cognitive impairment in cancer patients after completion of cancer therapy. Power calculations demonstrated that with our sample size of 34 subjects in each group, we would be able to detect an effect size of 0.62 standard deviations, using a 2-sided P value of <.05 for statistical significance. If the standard deviation of the change score was 1, then the difference of change score detected was 0.62; if the standard deviation was 2, then the difference detected was 2 × 0.62, with a power of 0.939. The primary independent variable was treatment group, and the dependent variable was the change in scores after randomization (Week 4 to Week 8). The primary outcome measures chosen for these comparisons were the change in the factor scores of Speed of Memory and Continuity of Attention, with the rationale that those patients who experience “chemobrain,” self-report problems with their memory and concentration,52 and to investigate whether these outcomes may prove sensitive to modafinil. Differences in assessment scores using paired samples from baseline (Week 0) to randomization (Week 4) and baseline to post-treatment (Week 8) were also compared across the entire study sample. If the distribution of the differences between pairs was non-normally distributed, the Wilcoxon signed rank test was used to assess the change within each treatment group. All statistical tests were 2-sided and performed at the 5% level of significance. Data analysis was performed using SAS software version 9.1 (SAS Institute, Cary, NC).
RESULTS
Subject Characteristics
Of the 82 eligible patients, 6 dropped out before completing the open-label phase; 2 complaining of anxiety, 2 reporting headaches, and 1 reporting nausea dropped out during the first week (all were very slight with weights <49.5 kg), and a final patient found herself unexpectedly pregnant midway through the study and was withdrawn (Fig. 2). The characteristics of the 6 participants who did not complete all assessments during randomization were very similar to those who did complete the study. The 6 patients were white, and most had received some college education. Most adverse events were considered by the investigator to be mild or moderate in severity (defined as no or some limitation of usual activities), occurred with greatest frequency during the first 2 weeks of open-label therapy, and were self-limited.
Patients who completed study at each time-point.
Seventy-six breast cancer survivors completed the 1-month open-label study (age range, 33-83 years; median age, 54 years). These women were Caucasian (n = 74, 97%), were married (n = 51, 67%), and had at least some college education (n = 57, 75%); the majority (n = 45, 59%) were employed in professional occupations. All subjects had been treated with surgery and chemotherapy, with 65 (86%) also receiving radiation therapy.
Upon assessment at Week 4, an additional 8 patients failed to meet the eligibility criteria to continue onto the randomization phase of the study, because they showed no improvement in fatigue during the open-label phase. The remaining 68 (83%) survivors were randomized, and they all fully completed the assessments—open-label and randomization—between Weeks 0 and 8 of the study. These 68 are included in these analyses. Baseline descriptive data for both the open-label phase and drug and placebo groups after randomization are shown in Table 2.
| Characteristics | Open-Label, Weeks 0-4, n=76 (%) | Modafinil, Weeks 4-8, n=34 (%) | Placebo, Weeks 4-8, n=34 (%) |
|---|---|---|---|
| Months since completed treatment | |||
| Mean [SD] | 22.8 [21.1] | 22.1 [22.7] | 22.1 [22.7] |
| Range | 1-96 | 1-96 | 1-96 |
| Age, y | |||
| Mean [SD] | 54.0 [10.3] | 51.76 [8.6] | 56.35 [11.4] |
| Range | 33-83 | 33-70 | 36-83 |
| Sex | |||
| Women | 76 (100) | 34 (100) | 34 (100) |
| Ethnicity | |||
| White | 74 (97.4) | 34 (100) | 33 (97) |
| Hispanic | 1 (1.3) | 0 (0) | |
| Black | 1 (1.3) | 0 (0) | 1 (3) |
| Education | |||
| Completed graduate training | 30 (39.5) | 14 (41) | 12 (35) |
| Standard college or university, 4 y | 20 (26.3) | 6 (17) | 12 (35) |
| Partial college training, <4 y | 7 (9.2) | 7 (21) | 0 (0) |
| High school graduate | 19 (25) | 7 (21) | 10 (30) |
| Marital status | |||
| Married | 51 (67.1) | 23 (67) | 23 (67) |
| Divorced | 6 (7.9) | 1 (3) | 4 (12) |
| Separated | 1 (1.3) | 0 (0) | 1 (3) |
| Single | 11 (14.5) | 8 (24) | 2 (6) |
| Widowed | 7 (9.2) | 2 (6) | 4 (12) |
| Previous surgery | 76 (100) | 34 (100) | 34 (100) |
| Previous chemotherapy | 76 (100) | 34 (100) | 34 (100) |
| Previous RT | 65 (85.5) | 28 (82) | 29 (85) |
| Other | 43 (56.6) | 18 (53) | 21 (62) |
- RT indicates radiotherapy.
The connection between aging and cancer is significant; >60% of all cancers occur among those >65 years old. There is little information regarding the effect of adjuvant chemotherapy on the older patient's cognitive abilities and whether the chemotherapy affects the patient's ability to maintain functional independence.53 The modafinil group mean age was 5 years younger than the placebo group in this analysis; however, independent t test revealed no statistically significant differences in age (P = .066), education (P = .921), or occupation (P = .182) between the 2 randomized groups modafinil (n = 34) and placebo (n = 34). Finally, we examined differences in fatigue and depression between the groups. In previous research, both fatigue and depression have demonstrated a potential to affect cognitive performance. However, no statistically significant differences between drug and placebo groups were observed for these variables (fatigue, P = .21; depression, P = .46).
Analyses
CDR battery
Overall results of the CDR composite mean scores are summarized in Table 3.
| Open Label, n=76, Week 0-4 | Modafinil, n=34, Week 4-8 | Placebo, n=34, Week 4-8 | |
|---|---|---|---|
| Speed of Memory† | |||
| Week 0 | 4543.22 (1165.4)1,2,3 | — | — |
| Week 4 | — | 4439.69 (1071.92)1 | 4314.74 (1020.17) |
| Week 8 | — | 4232.10 (1031.1)2,4 | 4335.56 (915.82)3,4 |
| Continuity of Attention‡ | |||
| Week 0 | 90.97 (3.79)5,6 | — | — |
| Week 4 | — | 92.00 (3.12) | 90.94 (3.12) |
| Week 8 | — | 92.24(2.09)5 | 91.59 (3.26)6 |
| Quality of Episodic Memory† | |||
| Week 0 | 211.49 (49.65)7,8 | — | — |
| Week 4 | — | 237.37 (40.53)7,9 | 232.20 (52.99) |
| Week 8 | — | 219.65 (52.50)8,9,10 | 224.17 (54.02)10 |
| Quality of Working Memorya | |||
| Week 0 | 1.82 (0.25) | — | — |
| Week 4 | — | 1.79 (0.38) | 1.80 (0.43) |
| Week 8 | — | 1.82 (0.20) | 1.70 (0.49) |
| Power of Attention‡ | |||
| Week 0 | 1275.08 (208.8) | — | — |
| Week 4 | — | 1228.24 (126.80) | 1276.33 (153.89) |
| Week 8 | — | 1247.80 (139.04) | 1275.44 (121.31) |
- Identical superscript numbers indicate a statistically significant difference between groups (P ≤ .05).
- CDR indicates cognitive drug research.
- * Mean (standard deviation).
- † Lower score is better.
- ‡ Higher score is better.
Open-label (Week 0 to Week 4)
In this study of 68 participants, the analysis of the primary outcome factor, the Speed of Memory Index, revealed a significant effect of modafinil (mean change, 240.003 seconds; P = .0073). This index was computed as a combination of 4 factors, which included speed of numeric working memory, speed of spatial working memory, speed of word recognition, and speed of picture recognition (Fig. 3). There was also a statistically significant effect of modafinil on the Digit Vigilance test (1 of the measures in Continuity of Attention; mean change, 1.623 U; P = .0014) during the 4 weeks of modafinil therapy. However, modafinil produced no significant effect on the composite Continuity of Attention (P = .0568).
Mean change from baseline is shown for Quality of Episodic Memory and Speed of Memory during the open-label phase (n = 76).
Three measures—immediate word recall (mean change, 8.59375; P < .001), word recognition accuracy (mean change, 6.985 U; P = .0002), picture recognition accuracy (mean change, −138.281 seconds; P = .0054)—in Quality of Episodic Secondary Memory also showed a statistically significant mean improvement from baseline to Week 4. Overall, significant improvement was observed in Quality of Episodic Secondary Memory (mean change, 26.826 U; P < .0001) from baseline to Week 4 (Fig. 3).
Modafinil showed no effect on Quality of Working memory (P = .2475) nor on Power of Attention (P = .9031) in the open-label phase.
Randomization (drug vs placebo, Week 4 to Week 8)
Randomization did not have a significant effect on each outcome measure of any 1-factor score. However, modafinil demonstrated a cumulative improvement from baseline to Week 8 in composite Speed of Memory scores (P = .029) compared with placebo (P = .116), and for composite Quality of Episodic Secondary Memory, the scores in the drug group show significant change (P = .015) compared with the placebo group (P = .155). Improvement was observed at the Week 4 visit and was maintained throughout the study.
About 70% of the subjects in the active drug group had an improvement in Continuity of Attention from baseline to post-treatment (Week 8), compared with 52% in the placebo group, but the difference was not significant (P = .19). The active drug produced a better response, an improvement over the Week 4 assessment, where no significant effect was seen (Fig. 4). At the final visit (Week 8), mean Continuity of Attention improved for the modafinil group (P = .010) compared with placebo (P = .54).
Cumulative improvement in Continuity of Attention is shown (Week 0 to Week 8). Group 1 indicates modafinil; group 2, placebo.
There was no significant effect of modafinil on Quality of Working Memory or Power of Attention during the randomization phase.
Study Limitations
This secondary analysis has some limitations. The patients in this study tended to be more educated than the general population and were primarily Caucasian, making these results less generalizable to patients from minority groups or those of lower socioeconomic status. In addition, the original study recruited 82 breast cancer patients, but for the purpose of this study, we have only analyzed data from the 68 patients who completed all 3 CDR assessments. Although this reduction in patient numbers may have resulted in a biased subsample, it should be remembered that patients were not randomized if they failed to respond to open-label modafinil in terms of the primary outcome of the original study—ie, fatigue, not cognitive improvement. Furthermore, as Table 1 demonstrates, the 68 completers did not appear to differ from the 76 individuals who had no side effects, and the purpose of this secondary analysis is to determine whether sufficient evidence indicates that modafinil improves cognitive function to support further trials to demonstrate this result more conclusively. For a future study, we propose a longitudinal design whereby patients are recruited before chemotherapy and randomized to modafinil or placebo, and cognitive performance is evaluated before chemotherapy and again 3 and 6 months postchemotherapy. This strategy would directly establish the presence of cognitive deficits related to chemotherapy and determine whether modafinil alleviated those deficits.
DISCUSSION
Modafinil is a wakefulness- and alertness-enhancing agent currently approved for treatment of excessive daytime sleepiness associated with narcolepsy, obstructive sleep apnea/hypopnea syndrome, and shift-work sleep disorder. Modafinil belongs to a class of drugs called eugeroics (meaning “good arousal”). Its uniqueness lies in its ability to only “stimulate when stimulation is required.”54 As a result, the “highs and lows” associated with other stimulants, such as amphetamine, are absent with eugeroics. Modafinil appears to produce its effects via different neural mechanisms than conventional stimulants or other psychotropic drugs with similar effects on wakefulness and alertness, and it does not seem to stimulate dopamine transmission in the brain or produce sympathomimetic effects. Although the precise mechanism of action of modafinil is unknown, numerous human and animal studies have elucidated many of its characteristics. Modafinil has site-specific central nervous system activity55, 56 and acts on a specific subset of brain pathways that regulate sleep and wakefulness. However, it does not bind to many receptors that are normally involved in sleep/wake regulation, including those for norepinephrine, serotonin, dopamine, and γ-aminobutyric acid. It does not have the effects on the extrapyramidal motor system—restlessness, hyperactivity, and irritability—that conventional stimulants exhibit, suggesting that modafinil has the potential to act as a cognitive enhancer without the side effects produced by other stimulants.
This study of modafinil, the first to examine cognitive function in patients after treatment for cancer, has demonstrated that 200 mg of modafinil enhances Speed of Memory, making patients both more accurate and faster at retrieving information. Although modafinil had no significant effect at Week 4 on Continuity of Attention, those participants who continued modafinil showed an improvement at Week 8. Significant improvement was also seen in the Quality of Episodic Secondary Memory, indicating that patients treated with modafinil were better able to store, retain, and retrieve both verbal and pictorial information. Modafinil was thus able to improve cognitive performance (memory and attention) in chemotherapy-treated breast cancer survivors.
Most of the effects of modafinil on the composite score of memory and some measures in attention were shown in the first postbaseline visit (Week 4), were sustained throughout the study, and are similar to those previously noted in clinical studies of modafinil in other populations.15-17, 57 The findings of this study demonstrate that the duration of an open-label drug is an important factor that needs to be addressed when considering a pharmacological intervention for a symptom. By commencing all our patients on 200 mg daily of modafinil and then dividing them into 2 groups, where 1 group continued the medication and the other started taking an identical placebo, we were able to compare 4 weeks of modafinil treatment with 8 weeks.
Quality of life, as well as survival and potential for cure, is of paramount importance to cancer patients. Post-treatment survivorship issues have become a prominent concern in cancer survivors and patients living with cancer. Cancer therapy-related cognitive decline, along with fatigue, anxiety, and depression, is 1 of the most common symptoms affecting quality of life.58, 59 During the past decade, studies have been performed to define this treatment-related cognitive decline and to study its prevalence and importance. Unfortunately, the precise etiology remains obscure for most patients with this debilitating side effect. A better molecular understanding of this cognitive dysfunction will likely lead to the identification of therapeutic targets and ultimately better treatment.
Although larger studies are needed to confirm these findings, the results of our secondary analysis of a trial of modafinil suggest that it can considerably improve some cognitive function measures in breast cancer survivors.
Conflict of Interest Disclosures
The study was supported by a grant from Cephalon, Inc, and National Cancer Institute Grants U10-CA37420 and R25-CA102618.
Keith Wesnes is the sole shareholder of CDR, Ltd, which provided the computerized tests in this study but did not charge any fee. Keith Wesnes has no stock interests in Cephalon.
