Pilot Study of the Prospective Identification of Changes in Cognitive Function During Chemotherapy Treatment for Advanced Ovarian Cancer
Change in cognitive function is increasingly being recognized as an adverse outcome related to chemotherapy treatment. These changes need not be severe to impact patient functional ability and quality of life. The primary goal of this study was to determine if there is evidence of changes in the cognitive function domains of attention, processing speed, and response time among women with newly diagnosed advanced ovarian cancer who receive chemotherapy. Eligible patients were women diagnosed with stage III–IV epithelial ovarian or primary peritoneal cancer who had not yet received chemotherapy but who were prescribed a minimum of six cycles (courses) of chemotherapy treatment. Cognitive function was assessed by a computerized, Web-based assessment (attention, processing speed, and reaction time) and by patient self-report. Cognitive function was assessed at three time points: prior to the first course (baseline), course three, and course six. Medical records were reviewed to abstract information on chemotherapy treatment, concomitant medications, and blood test results (eg, hemoglobin, CA-125). Of the 27 eligible participants, 92% and 86% demonstrated cognitive impairments from baseline to course three and from baseline to course six of chemotherapy, respectively. Impairment was detected in two or more cognitive domains among 48% (12 of 25) and 41% (9 of 22) of participants at course three and course six of chemotherapy, respectively. This study shows evidence of decline in cognitive function among women being treated for ovarian cancer. There is a need for additional, prospective research to better understand the impact of chemotherapy on cognitive function among ovarian cancer patients so that effective preventive and treatment strategies can be developed.
| n = 27 | |
| Mean age, years (range) | 59.3 (40.3–81.5) |
| Education, n (%) | |
| High school or less | 3 (11.1%) |
| Some college | 12 (44.4%) |
| College graduate | 12 (44.4%) |
| Race/ethnicity, n (%) | |
| White, non-Hispanic | 25 (92.6%) |
| Hispanic | 1 (3.7%) |
| Native American | 1 (3.7%) |
| Marital status, n (%) | |
| Married/cohabitating | 19 (70.4%) |
| Divorced/separated | 1 (3.7%) |
| Widowed | 5 (18.5%) |
| Never married | 2 (7.4%) |
| Mean courses of chemotherapy, n (range) | 5.9 (4–6) |
| Chemotherapy route, n (%) | |
| Intraperitoneal | 5 (18.5%) |
| Intravenous | 22 (81.5%) |
| Concurrent medication use, n (%) | |
| Antidepressant | 7 (25.9%) |
| Antianxiety | 16 (59.3%) |
| Sleep aids | 5 (18.5%) |
Web-Assessed Cognitive Function
Keyboard proficiency remained unchanged over time (P = 0.39). As shown in Table 2, most participants demonstrated cognitive impairments in at least one of the three cognitive domains assessed during this study (92% and 86% at course 3 and course 6, respectively). Nearly half of the study participants demonstrated impairment from baseline in two or more of the three cognitive domains assessed (Table 3). Table 4 shows a detailed summary of the subscales within the Web-based cognitive tests that comprised the CIS.This table demonstrates the statistically significant increase in test subscale errors, despite the test-taking improvements over time, as shown by reduction in testing time.
| CIS | COURSE 3 | COURSE 6 |
|---|---|---|
| No decline (CIS = 0) | 2 (8%) | 3 (14%) |
| One impairment (CIS = 1) | 11 (44%) | 10 (45%) |
| Two impairments (CIS = 2) | 11 (44%) | 7 (32%) |
| Three impairments (CIS = 3) | 1 (4%) | 2 (9%) |
| COGNITIVE IMPAIRMENT SCALE (CIS) FACTORS | BASELINE | COURSE 3 | COURSE 6 | |||||||
|---|---|---|---|---|---|---|---|---|---|---|
| N | MEAN | SD | N | MEAN | SD | N | MEAN | SD | P | |
| Attention | ||||||||||
| Number recall (number correct) | 25 | 7.08 | 1.75 | 25 | 7.16 | 2.03 | 22 | 7.45 | 1.92 | 0.887 |
| Number sequencing (number correct) | 26 | 6.23 | 0.98 | 25 | 5.96 | 2.65 | 23 | 5.61 | 2.29 | 0.476 |
| Processing speed | ||||||||||
| Animal decoding (number of errors) | 25 | 0.4 | 0.5 | 25 | 0.72 | 0.84 | 23 | 3.26 | 0.86 | <0.0001 |
| Animal decoding (number correct) | 25 | 32.48 | 6.48 | 25 | 32.96 | 8.90 | 23 | 32.22 | 8.70 | 0.678 |
| Symbol scanning (number correct) | 27 | 18.59 | 1.15 | 25 | 18.76 | 1.2 | 21 | 18.67 | 1.35 | 0.883 |
| Symbol scanning (response time) | 27 | 4.38 | 1.37 | 25 | 4.26 | 1.66 | 21 | 3.61 | 0.84 | 0.002 |
| Reaction time | ||||||||||
| Response direction 1 (number of omissions) | 27 | 0.04 | 0.19 | 26 | 0.62 | 2.35 | 23 | 0 | 0 | 0.028 |
| Response direction 1 (response time, seconds) | 27 | 0.52 | 0.06 | 26 | 0.55 | 0.22 | 23 | 0.52 | 0.07 | 0.567 |
| Response direction 2 (number of omissions) | 27 | 0.63 | 1.33 | 26 | 0.5 | 2.18 | 23 | 0.43 | 0.95 | 0.135 |
| Response direction 2 (response time, seconds) | 27 | 0.75 | 0.13 | 26 | 0.72 | 0.20 | 23 | 0.71 | 0.17 | 0.467 |
| Response direction, shift failures (number) | 27 | 4.33 | 3.13 | 26 | 2.77 | 2.29 | 23 | 3.04 | 2.58 | 0.007 |
The mean values and 95% confidence intervals of the patient-reported cognitive function outcomes are presented in Figure 1. Mean values remained within the low impairment range (less than 1.25) during chemotherapy.
Blood Chemistries and Toxicity
The mean values and 95% confidence intervals of significant differences in blood chemistries and toxicities are presented in [Figure 2] and [Figure 3]. Total patient-reported neurotoxicity increased significantly during chemotherapy (ANOVA; F = 6.851, P = 0.002), while several mean blood chemistry values decreased during chemotherapy treatment (hemoglobin F = 2.465, P = 0.09; white blood cell count F = 16.95, P < 0.001; platelets F = 13.72, P < 0.001; and CA-125 F = 4.91, P = 0.01). One study participant received a blood transfusion at the final course of chemotherapy, and two and three participants received cytokines (erythropoietin or darbepoietin) at course 3 and course 6, respectively.
Discussion
This study shows preliminary evidence that cognitive decline is a significant factor experienced by women who are treated for advanced ovarian cancer. Most participants self-reported mild declines, and these were detectable by a sensitive Web-based assessment tool. There are many potential mechanisms of cognitive decline during chemotherapy, ranging from oxidative damage to reduced blood oxygenation due to anemia to stress and anxiety. While it is outside of the scope of this small pilot study to examine the causative factors of decline, it does suggest the need for further investigation of the effect and potential mechanisms of cognitive decline in this population. While most of the prior work in cognitive function has been conducted among breast cancer patients, ovarian cancer patients appear to experience cognitive decline as well. There is a need to further understand this issue so that effective preventive or treatment strategies can be developed.
The significant increase in patient-reported neurotoxicity across each study visit may be a concern for computerized assessments that require dexterity. However, the keyboard proficiency tests did not decline over time, suggesting that the neurotoxicity reported by patients in this study was not great enough to affect their ability to use the computer keyboard. Patients appear to report higher levels of difficulty with memory (eg, forgetfulness) following diagnosis than following the initiation of chemotherapy; however, higher-level cognitive processes (eg, logic, organizational abilities, calculations) reported by patients appear to decline following the initiation of chemotherapy. Although larger, adequately powered trials are needed to determine the extent of this decline, this suggests that patients experience increasing challenges that may interfere with their ability to perform necessary tasks at work and in the household. Further work is needed to examine the duration of these effects following chemotherapy. Since the cognitive impact of chemotherapy reported by patients is mild, investigators must ensure the use of appropriate patient-reported tools that are able to detect these differences. While reported decline may occur, this is likely to remain within the mild category of traditional assessment tools. It is of benefit to use patient-reported tools such as the PAF that also permit the analysis of continuous data.
This study is limited by its design as a pilot study and was challenged by several logistical issues. Four patients were unable to complete all the neurocognitive evaluations. This was due to remote study staff, who would visit various clinics in the Tucson and Phoenix metropolitan regions in Arizona (range of travel more than 120 miles). The lack of completion was entirely due to communication and travel complications. When a patient was rescheduled to a different chemotherapy date, it was not always possible for this to be communicated to the Arizona Cancer Center researchers in a timely manner, resulting in missed visits. It is recommended for future studies that require strict timelines for study assessments (such as this cognitive function study) that the assessments be conducted by staff in those practices who can identify changes in infusion dates when they occur. This will reduce the communication barriers and rate of missed visits. This study was also not designed to be a comprehensive assessment of neurocognitive function but was focused on assessing three domains: attention, processing speed, and response time. It is possible that many other domains of cognitive function could be impacted by chemotherapy that were not evaluated in this study. Many patients were also taking antidepressant medications during the study; however, these were generally not new prescriptions and were also being taken at the baseline assessment. Nevertheless, future studies should incorporate assessments of mood, depression, and anxiety to account for the potential effect of these factors on cognitive assessment scores.
Despite these limitations, the study provides preliminary data demonstrating cognitive decline during chemotherapy among ovarian cancer patients treated in the front-line setting of advanced disease. More than 90% of all patients experienced measurable impairments in cognitive function during primary chemotherapy. More than half of all patients demonstrated impairment on two or more cognitive domains. Prior work has shown that even mild cognitive impairments can influence quality of life and the ability to perform routine daily activities (eg, taking medications, returning to work, managing household finances).23 The data emphasize the critical need to further understand the impact of chemotherapy on cognitive function among ovarian cancer patients so that effective preventive and treatment strategies can be developed. Additional research is needed to understand how long these declines may persist following chemotherapy treatment.
