Changes in Attentional Function in Patients From Prior to Through 12 Months After Breast Cancer Surgery

INTRODUCTION

Cognition is a fundamental aspect of one’s ability to engage in activities and accomplish goals. Cognition involves different inter-related functions, including attention, memory, comprehension, problem solving, verbal fluency, and decision-making.¹ Directed attention is connected to working memory and requires mental effort to actively inhibit irrelevant distractions for goal-directed tasks. Attentional function, supported by directed attention, may be among the most essential functions for cancer patients because it helps them acquire important information, make decisions, and carry out self-care-activities.² Impairments in attentional function may interfere with patients’ self-care activities, such as their ability to adhere to treatment, manage side effects, and re-integrate into the workforce, which has a negative impact on their quality of life (QOL).^3,4

Approximately 75% of breast cancer patients experience cancer-related cognitive impairment (CRCI) prior to, during, or after treatment.⁵ In addition, CRCI remains a significant long-term problem for about 35% of survivors.^6-11 Determining the underlying mechanisms and risk factors for, as well as effective interventions for CRCI has been challenging. Findings from objective neuropsychological tests correlate poorly with self-report measures of CRCI,^2,12-16 and often do not detect deficits, even when patients have subjective complaints about difficulties with routine activities (e.g., problems with driving, inability to return to work).^13,17 Therefore, self-report measures are an essential component of the assessment of CRCI because they are better able to capture patients’ experiences.^18,19

Initially, CRCI was hypothesized to occur only with the administration of chemotherapy (CTX; i.e., “chemobrain”). However, accumulating evidence suggests that the etiology of CRCI in breast cancer patients is multi-factorial (e.g., the cancer itself,⁸ neurotoxic effects of CTX;^20-26 chemically or surgically-induced menopause and the effects of hormone treatments on estrogen levels;^4,27-29 systemic immune responses induced by localized radiation therapy and subsequent elevation in plasma interleukin (IL)-6 levels;^30,31 delirium and stress associated with anesthesia and sugery;³² chronic inflammation, dysregulation of cytokines, and premature aging due to telomere shortening induced by the cancer itself^7,32,33) Because most of these data come from studies that evaluated CRCI following surgery,³³ additional research is warranted that addresses the occurrence of and risk factors for CRCI prior to and following breast cancer surgery.

A number of demographic and clinical characteristics, co-occurring symptoms, and psychosocial adjustment characteristics are associated with an increased risk for more severe CRCI. For example, while older breast cancer patients are presumed to be at a higher risk for CRCI because of age-related changes in cognition and the use of different coping mechanisms,² in some studies,^23,33-35 younger patients were at increased risk for CRCI during the survivorship period. Co-occurring symptoms, such as anxiety,^9,36,37 depressive symptoms,^6,36,37 fatigue,^6,8,9,37,38 neuropathy,³⁹ and sleep disturbance,^6,37,38 have been associated with CRCI both prior to and after breast cancer surgery.¹³ In addition, in studies that evaluated patients prior to breast cancer surgery,^2,6,40 while CRCI was not associated with treatment modality, it was associated with pre-treatment levels of co-occurring symptoms. Of note, findings from recent studies suggest that associations between CRCI and anxiety and depressive symptoms may be related to challenges with emotional regulation, coping (e.g., avoidance), and psychosocial adjustment to breast cancer.^41-44 However, because of the observational nature of these studies, the relatively small sample sizes, and the lack of a comprehensive assessment of risk factors, these findings warrant replication to obtain a more comprehensive understanding of the risk factors associated with CRCI prior to and following breast cancer surgery.

In several longitudinal studies that evaluated for CRCI in women with breast cancer^{6,8,9,11,36,45} findings suggest that cognitive function returns to presurgical levels at approximately 12 months. However, direct comparisons across studies are difficult because of the different aspects of CRCI that were evaluated (e.g., memory, verbal fluency); differences in the use of subjective versus objective measures; and differences in the number and timing of the assessments. In the one study that evaluated for inter-individual differences in CRCI trajectories,⁶ the patients’ cancer diagnosis and/or co-occurring symptoms were associated with CRCI at each measurement point, from prior to through 24 months post-surgery. Of note, two recent reviews highlighted the paucity of research in this area and recommended that longitudinal studies examine the associations between CRCI and common co-occurring symptoms.^36,42

Taken together these findings suggest that breast cancer patients are at relatively high risk for CRCI and that additional research is warranted on inter-individual differences in risk factors for this significant clinical problem. Given that no study has evaluated a comprehensive list of potential risk factors for CRCI from prior to through recovery after surgery, the purposes of this study, in a sample of patients who were assessed from prior to through 12 months after surgery (n=396), were to evaluate for inter-individual variability in the trajectories of self-reported attentional function and to determine which demographic, clinical, symptom and psychosocial adjustment characteristics were associated with initial levels and with changes in the trajectories of attentional function.

METHODS

Statistical analyses

Descriptive statistics and frequency distributions were generated on the sample characteristics, symptom severity scores, and psychosocial adjustment items using SPSS version 23.⁶⁸ AFI was assessed prior to surgery and at 1, 2, 3, 4, 5, 6, 8, 10, and 12 months after surgery. All other demographic, clinical, symptom and psychosocial adjustment characteristics that were evaluated as predictors in the hierarchical linear modeling (HLM) analysis were assessed prior to surgery.

HLM based on full maximum likelihood estimation was done using the software developed by Raudenbush and colleagues.^69,70 This analysis is discussed in detail in our previous publications.^61,71,72 In brief, during stage 1, intra-individual variability in AFI scores over time was examined. At this point, the model was constrained to be unconditional (i.e., no predictors) and likelihood ratio tests were used to determine the best fitting model.

The second stage of the HLM analysis examined inter-individual differences in the trajectories of AFI scores by modeling the individual change parameters (i.e., intercept, linear slope) as a function of proposed predictors at level 2. Table 1 presents a list of the proposed predictors that was developed based on a literature review about CRCI in breast cancer patients.^{6,16,40,44,67,73,74} To improve estimation efficiency and construct a parsimonious model, an exploratory level 2 analysis was completed in which each potential predictor was assessed to determine whether it would result in a better model if it alone were added as a level 2 predictor. Predictors with a t-value of < 2.0 were dropped from subsequent model testing. All potential significant predictors from the exploratory analyses were entered into the model to predict each individual change parameter. Only predictors that maintained a statistically significant contribution in conjunction with other variables were retained in the final model. A p-value of <.05 indicated statistical significance.

Table 1 –

Potential predictors of the intercept (I) and linear coefficient (LC) for attentional function using preoperative characteristics

Characteristic	I	LC
Demographic characteristics
Age
Lives alone
Education
Marital status
Ethnicity
Employment status
Clinical characteristics
Body mass index
Karnofsky Performance Status score
Self-administered Comorbidity Questionnaire score
Neoadjuvant chemotherapy
Type of surgery
Sentinel lymph node biopsy
Axillary lymph node dissection
Breast reconstruction at the time of surgery
Menopausal status
Adjuvant radiation therapy in the first six months after surgery
Adjuvant chemotherapy in the first six months after surgery
Use of complementary therapy in first six months after surgery
Use of hormonal therapy in the first six months after surgery
Symptoms
Trait anxiety score
State anxiety score
Center for Epidemiological Studies Depression Scale score
General Sleep Disturbance Scale score
Lee Fatigue Scale score
Lee Energy Scale score
Attentional Function Index score
Presence of breast pain prior to surgery
Worst pain
Average pain
Number of days per week in pain
Number of hours per day in pain
Severity of hot flashes
Psychosocial adjustment characteristics
Distress at initial cancer diagnosis
Fear of future diagnostic tests
Fear of metastasis
Difficulty coping as a result of disease/treatment
Control of things in your life

Open in a new tab

RESULTS

DISCUSSION

This study is the first to use HLM to examine inter-individual differences in the trajectories of attentional function and to evaluate whether a comprehensive list of demographic, clinical, symptom, and psychosocial adjustment characteristics were associated with AFI scores prior to surgery and over a period of 12 months. Prior to surgery, our patients had a mean AFI score of 6.5 which suggests that these women had some pre-existing impairments in attentional function. Our finding is consistent with a study by Cimprich and collegues (i.e., preoperative AFI score of 6.6)² but lower than the AFI score reported by Chen and colleagues (preoperative AFI score of 8.2).⁶ While our sample’s overall AFI score improved over the twelve months following surgery, at the end of one year, it remained in the moderate range (i.e., estimated at 7.18). However, as seen in Figure 3D, and consistent with a previous report,⁷⁵ patients in our study with lower preoperative AFI scores demonstrated greater improvements in attentional function in the 12 months following surgery. The minimal increase in AFI scores in our sample may be associated with inter-individual variability in the relatively large number of characteristics associated with decrements in attentional function that were identified in this study.

Consistent with previous reports,^9,36 no differences in preoperative AFI scores were found between our White and non-White patients. However, over the 12 months of the study, compared to the White patients, non-White patients had less improvements in their AFI scores over time. This association may be related to a number of factors including level of education,^76,77 employment status and ability to return to work,⁷⁸ and/or financial burden and high out-of-pocket costs associated with cancer and its treatments.⁷⁹ Our findings suggest that breast cancer patients from diverse ethnic backgrounds may need additional support and resources to improve their level of attentional function.

In terms of clinical characteristics and consistent with previous reports,^80-82 a higher level of comorbitiy was associated with lower preoperative AFI scores. In addition, while having an axillary lymph node dissection was not associated with changes in attentional function, patients who did not have a SLNB (i.e., 17.4% of the sample) reported worse AFI scores over time. This finding is most likely related to the fact that women who did not have a SLNB were most likely to be diagnosed with more advanced stage disease and had received neoadjuvant CTX and/or a mastectomy with an axillary lymph node dissection.

While receipt of neoadjuvant CTX was identified as a potential predictor in the exploratory analyses, it was not included in the final model. However, consistent with previous reports,^23,36,75,83 receipt of adjuvant CTX was associated with lower AFI scores. In addition, patients who received hormonal treatment had a decline in attentional function over the 12 months of the study. The molecular mechanisms associated with CTX- and hormonally-induced changes in cognitive function are extremely complex and include: disruption of the blood-brain-barrier; deoxyribonucleic acid (DNA) damage and associated deficits in DNA repair mechanisms; telomere shortening; oxidative stress and associated inflammatory responses; polymorphisms in genes associated with neural repair; alterations in neural transmission; as well as changes in levels of sex steroid hormones that under physiologic conditions are neuroprotective and reduce oxidative stress (for reviews see 84-86). Additional research is warranted that examines associations between changes in attentional function and each of these molecular mechanisms. These types of investigations may identify potential therapeutic targets.

This study is the first to evaluate for associations between a comprehensive list of common co-occurring symptoms and changes in attentional function. Of note, in the exploratory analysis, all of the symptoms evaluated (i.e., trait anxiety, state anxiety, depression, fatigue, decrements in energy, sleep disturbance, pain, hot flashes) were associated with initial levels and/or the trajectories of attentional function. However, only trait anxiety, fatigue, decrements in energy, and sleep disturbance were retained in the final model. Consistent with previous studies,^6,8,36,87 higher levels of fatigue and lower levels of energy were associated with poorer attentional function at enrollment that persisted over the 12 months of the study. The association among these three symptoms may be related to a number of shared pathophysiologic mechanisms including: increases in inflammatory responses,⁸⁸, as well as the deleterious effects of cancer^87,89 and its treatments^8,9,87 on DNA repair mechanisms and neural transmission.

Consistent with previous studies,^{2,6, 16,38,73,90} higher levels of sleep disturbance were associated with lower AFI scores. At enrollment, our patients reported relatively high levels of sleep disturbance (i.e., mean GSDS score 48.1 (α21.3); cut-off score of ≥43 indicates a significant level of sleep disturbance). Again, the association between decrements in attentional function and sleep disturbance may be due to decreases in the restorative processes associated with 6 to 8 hours of sleep^91,92 as well as excessive daytime sleepiness which is often associated with sleep disturbance. In addition, these two symptoms may share common underlying mechanisms including: increased levels of pro-inflammatory cytokines, changes in activity with the hypothalamic-pituitary adrenal axis, and alterations in DNA repair mechanisms.^33,87,93,94

Consistent with other studies,^{2,6,9,36,44,45,75,95,96} associations were found between higher preoperative levels of trait and state anxiety and lower levels of attentional function. Our patients reported levels of trait and state anxiety that were above the clinically meaningful cutoff scores for the instruments. In addition, findings from several longitudinal studies suggest that positive associations between anxiety and decrements in attentional function persist over time.^6,9,45 Of note, while anxiety and depressive symptoms often co-occur in cancer patients,⁹⁷ even though depressive symptoms were identified as a potential predictor in the exploratory analysis, it was not retained in the final model. Additional research is warranted on the co-occurrence of these two psychological symptoms and associated decrements in attentional function.

The relationship between higher levels of state anxiety and decrements in attentional function may be partially explained by women’s use of disengagement coping strategies (e.g., avoidance, denial, self-blame).^41,44,98 This hypothesis is partially supported by our finding that women who reported that they perceived that they had less control over things in their life had lower levels of attentional function at enrollment. However, it is not readily apparent why an association was found with lower levels of difficulty with coping at enrollment and lack of improvements in attentional function over time. One could hypothesize that patients with more difficulties coping prior to surgery were more likely to seek and obtain support to improve their coping skills and had associated improvements in cognitive function. That said, our findings suggest that clinicians need to assess patients prior to surgery for anxiety, as well as various psychosocial adjustment characteristics (e.g., coping styles) so that patient education and/or appropriate referrals can be initiated.^99,100

While this longitudinal study had numerous strengths including a relatively large sample size; a comprehensive evaluation of associations between demographic, clinical, symptom, and psychosocial adjustment characteristics and decrements in self-reported attentional function; as well as 12 months of follow-up, several limitations warrant consideration. The majority of our patients were White, well educated, and diagnosed with early stage disease, which limits the generalizability of our study findings. Despite the fact that findings from objective neuropsychological tests correlate poorly with self-report measures of CRCI,^2,12-16 and often do not detect deficits, even when patients have subjective complaints about difficulties with routine activities,^13,17 future studies should evaluate for longitudinal changes in CRCI using both types of measures. In addition, future studies should include age-matched controls without breast cancer. Our analysis focused on an evaluation of associations with preoperative levels of co-occurring symptoms. Given that symptom severity changes over time, future studies need to consider the use of statistical techniques like parallel process growth modeling to evaluate for changes over time in the severity of attentional function and each of the co-occurring symptoms.

While the clinical characteristics associated with deficits in attentional function (i.e., level of comorbidity, receipt of adjuvant CTX, receipt of hormonal therapy) cannot be changed, many of the other characteristics can be assessed and addressed with appropriate pharmacologic and non-pharmacologic interventions. The initiation of these interventions, in the perioperative period, may result in a decrease in patients’ overall symptom burden, as well as improvements in attentional function. In addition, timely referrals to mental health professionals and/or social workers may decrease patients’ levels of anxiety and stress, as well as improve their ability to cope with their cancer treatment(s) and maintain their level of cognitive function.