Cognition and Return to Work Status 2 Years After Breast Cancer Diagnosis

Key Points Question Is cognition associated with return to work 2 years after breast cancer diagnosis? Findings In this case series of 178 individuals with breast cancer in the French Cancer Toxicities trial cohort, return to work 2 years after breast cancer diagnosis was associated with higher cognitive speed performances before and after cancer treatment. Meaning These findings suggest that cognitive difficulties among individuals treated for breast cancer should be assessed before return to work to propose suitable management.


Introduction
Due to improvement in the survival rate, employment after cancer treatment is an important issue for individuals with breast cancer (BC).Many women wish to return to work (RTW) after BC treatment in order to lead an active socioprofessional life, enjoy a better quality of life, and have financial security.[3][4] Cancer-related cognitive impairment (CRCI) is frequently reported by patients with BC: 50% or more report cognitive difficulties after adjuvant chemotherapy and 15% to 25% have objective cognitive decline. 5,6Cancer-related cognitive impairment mainly concerns processing speed, executive function, memory and attention difficulties. 7,8It can be assessed with self-reported measures (cognitive symptoms) or with cognitive tests (objective cognitive functioning).Systematic reviews have reported that cognitive symptoms are not associated with objective cognitive functioning but are frequently associated with anxiety, depression, and fatigue. 9,10ncer-related cognitive impairment may be associated with difficulties with RTW.For example, 76% of patients with cognitive symptoms reported difficulties with RTW. 11[14] A systematic review reported that most studies have focused on the association between cognitive symptoms and work-related outcomes, such as work productivity and ability. 15Thus, higher levels of cognitive symptoms and fatigue 2 to 10 years after BC diagnosis were associated with lower future work ability. 16In persons with BC, cognitive symptoms 5 years after treatment were associated with poorer work ability, performance, and productivity. 17w studies have assessed the association between objective cognitive functioning and workrelated outcomes, 15 and, to our knowledge, all of them were cross-sectional.Among them, a small case-control study observed that more individuals with cancer who had cognitive impairment (7 of 15) did not RTW 1 year after the first day of breast cancer-related sick leave in comparison with individuals without impairment (9 of 30), but the small sample size limited the results. 18In a larger sample of individuals with cancer who reported cognitive symptoms, small effect sizes were found for the association between work-related outcomes assessed 3 years after cancer diagnosis (work ability and physical functioning at work) and objective cognitive performances (mainly, overall cognition and motor performance). 19ong the studies on cognitive symptoms and work-related outcomes, to our knowledge, none assessed cognition before RTW.Furthermore, only 1 small study assessed the association between objective cognition and work status without cognitive assessment before RTW. 18Better understanding of the association between cognition and work status could allow vocational rehabilitation to be offered to individuals with CRCI after BC treatment, thereby preventing or minimizing the work-related difficulties associated with CRCI.
The main aim of this longitudinal case series study based on a large nationwide cohort was to examine whether cognition, assessed using objective and subjective scores, was associated with RTW 2 years after BC diagnosis in patients who were employed or looking for employment at diagnosis.Factors such as anxiety, depression, and fatigue were also considered.We hypothesized that cognitive functioning, assessed before RTW, would be associated with RTW.

Participants and Study Design
Data for this case series were obtained from a substudy of the nationwide French Cancer Toxicities (CANTO) cohort (n = 12 000) of women with stage I to III BC 20 21 For the purpose of this study, we selected women who were working or looking for a job at diagnosis, aged 58 years or younger (to allow a 5-year delay before they reached the minimum retirement age) at the time of BC diagnosis and who had work-related characteristics at year 2. Furthermore, we selected only women who had undergone cognitive assessment at baseline and year 2 and for whom age data were available.The study was approved by the French regulatory authorities (Comité de protection des personnes Ile de France VII).All participants provided written informed consent and received financial compensation.

Outcome and Measures
The outcome was RTW assessed 2 years after BC diagnosis.Return to work was dichotomized as RTW or no RTW (including sick leave, looking for a job, or disability status).
We investigated the following exposure variables.

Statistical Analysis
Baseline characteristics, cognitive measures, and PROs were described and compared according to RTW at year 2 by the t test for quantitative variables and the χ 2 or Fisher exact test, if necessary, for qualitative variables.Two-sided hypothesis tests were constructed with a 5% level of significance.
Multivariable logistic regression models were used to explain RTW at year 2 according to cognitive measures and PROs at year 2 (cross-sectional association), as well as PROs at baseline and year 1 (longitudinal association).One model per cognitive measure and/or PRO was constructed, each adjusted on age, occupational class (2 classes), cancer stage, and chemotherapy.Results of logistic regression models provided odds ratios (ORs) with 95% CIs and P values, with ORs greater than 1 indicating a positive association with RTW.For quantitative variables, ORs are presented with their corresponding unit of increase (eg, 1-pt OR for an increase of 1 point).Missing PRO data at follow-up were imputed using the scores obtained at the previous assessment. 22As there were few other missing data (<10%) (Table 1), 29

Demographic and Clinical Characteristics
Among
Patients with RTW status were significantly less likely to have a mastectomy (24.1% vs 54.1%; P < .001)and had cancer greater than or equal to stage II less often (61.7%vs 81.1%; P = .04).There was no significant difference in other clinical data or treatment between groups.
There was no significant difference between groups for cognitive fatigue, anxiety, and depression.

Discussion
To our knowledge, this longitudinal case series is the first using a prospective cohort to examine whether cognition, including objective and subjective scores, is associated with RTW 2 years after BC diagnosis.Processing speed and executive function performance at diagnosis and after treatment completion were the main objective cognitive domains associated with RTW, as was physical fatigue.
Two years after BC diagnosis, lower overall cognitive impairment; higher performance of working memory, processing speed, and attention; higher perceived cognitive abilities and lower depression were associated with RTW.
In this substudy of the CANTO study, we found that 20.8% of patients with BC who were working (n = 170) or looking for a job (n = 8) before cancer treatment had not returned to work 2 years after BC diagnosis.This is in line with the proportion found in the whole CANTO cohort 1 and similar to that of a French national population-based survey. 30 found previously, RTW was associated with several factors, including socioeconomic (occupational class) and clinical (type of surgery) characteristics, working conditions (strenuous work, shift work, and low independence in decision-making), and psychological factors (depression). 31Furthermore, in line with previous reports, RTW at year 2 was associated with persistent cancer treatment-related adverse effects, such as physical fatigue, assessed before treatment and at year 1. 32 To our knowledge, this case series is the first longitudinal study to observe that objective cognitive functioning is also associated with RTW status 2 years after BC diagnosis.Processing speed and executive function, assessed at diagnosis before cancer treatment and 1 year after cancer diagnosis, were associated with RTW at year 2. Thus, before treatment and after adjuvant treatment, patients with BC who had impaired cognitive speed and/or executive function, such as inhibition or flexibility, were those who had lower RTW status 2 years after diagnosis.
In addition, we found cross-sectional associations between work activity at year 2 and overall cognitive impairment, working memory, processing speed, and attention, perceived cognitive abilities, and depression.Cognitive symptoms assessed at baseline and at year 1 were not associated with RTW status at year 2.However, they were associated when they were assessed at the same time as RTW, such as at year 2. Previous studies have also observed that cognitive symptoms are higher among patients who do not RTW, 32 but to our knowledge, no study has assessed cognitive symptoms before RTW.Processing speed was associated with RTW status at year 2, regardless of assessment time.
Studies have previously reported that this cognitive domain is frequently associated with RTW, particularly in patients with meningioma or after traumatic brain injury. 33,34 observed that more patients with objective cognitive impairment (35.1% [13 of 37]) at year 2 than without (14.2%[20 of 141]) did not RTW.This finding is consistent with a previous small study in which more individuals with cancer with (46.7% [7 of 15]) than without (30.0%[9 of 30]) cognitive impairment did not RTW. 18Nevertheless, this finding was not significant in that study, whereas the OR with adjustment was significant in ours.
17]35 We found that objective cognitive domains, such as processing speed and executive function, assessed at diagnosis and the end of treatment were associated with RTW at year 2. As RTW after cancer is an important issue for individuals with BC, management of CRCI could be offered to patients whose processing speed and executive function have been impaired.Management for CRCI could be proposed by the occupational physician during sick leave or when RTW is planned.This timing issue requires further investigation, since a CRCI intervention before cancer treatment has been completed could compromise optimal management according to fatigue or other adverse effects induced by cancer treatment.Cognitive training has been shown to improve CRCI 36,37 and experts recommend combining it with psychoeducation as part of a cognitive rehabilitation program. 38Furthermore, specific vocational rehabilitation, including training to increase processing speed and executive function, could be developed for women with cognitive difficulties who plan to RTW after BC treatments.b Missing data at year 1, n = 7 of 178 (3.9%).
CANTO-Cog patients (n = 494; mean [SD] age, 53.7[10.5]years;cancerstageՆII:270 of 492 [54.9%]), 299 had a baseline cognitive assessment, were aged 58 years or younger, and were working or looking for a job at diagnosis(Figure).Of these, 51 had no year 2 cognitive assessment and 70 had unknown employment status at year 2. The final sample included 178 women with BC (median age, 48.7 [range,28-58]years; there were no losses to follow-up), including 170 women who were working and 8 patients looking for a job at baseline.Of these, 141 were at RTW status at year 2.Among 37 women (20.8%) who did not RTW at year 2 (Table1), 24 were on sick leave, 6 were looking for a job, 6 were disabled, and 1 had another status.Among the 8 patients looking for a job at baseline, only 1 was looking for a job at year 2. At baseline, 35 of all patients (19.6%) had overall cognitive impairment.

Table 1 .
Baseline Demographic and Clinical Characteristics and Year 2 Work-Life Imbalance, Work Characteristics, and Overall Objective Cognitive Impairment (continued) -pt OR, 2.88; 95% CI, 1.36-6.28;P= .006),andlower physical fatigue (10-pt OR, 0.84; 95% CI, 0.71-0.98;P= .02).Altogether, RTW status at year 2 was associated with processing speed performance at each time.Return to work at year 2 was associated with processing speed and executive performance and physical fatigue assessed at baseline and year 1.Return to work at year 2 was not associated with cognitive symptoms assessed at baseline or at year 1.
a Clerks, self-employed, manual workers, and farmers.b No. of time per months where work was performed at night for at least 2 hours, between 10 PM and 5 AM.c Level 3 on the World Health Organization analgesic ladder, anxiolytics, antidepressant treatments, and hypnotics.JAMA Network Open | Oncology Cognitive and Return to Work Status After Breast Cancer Diagnosis JAMA Network Open.2024;7(8):e2427576.doi:10.1001/jamanetworkopen.2024.27576(Reprinted) August 19, 2024 6/13 Downloaded from jamanetwork.comby guest on 09/19/2024 performance (1

Table 2 .
Baseline Cognition and Patient-Reported Outcomes According to RTW Status Abbreviations: PCA, perceived cognitive abilities; PCI, perceived cognitive impairment; QOL, quality of life; RTW, return to work.

Table 3 .
Year 1 and Year 2 Cognition and Patient-Reported Outcomes According to RTW Status Abbreviations: PCA, perceived cognitive abilities; PCI, perceived cognitive impairment; QOL, quality of life; RTW, return to work.aissing data at year 1, n = 7 of 178 (3.9%).

Table 4 .
Cognition Scores and Patient-Reported Outcomes Associated With RTW Status at Year 2 Abbreviations: OR, odds ratio; PCA, perceived cognitive abilities; PCI, perceived cognitive impairment; QOL, quality of life; RTW, return to work.aEach measure corresponds to separate multivariable models adjusted for age, occupational class, cancer stage at diagnosis, and chemotherapy.For quantitative variables, ORs are presented with their corresponding unit of increase (eg, 1-pt OR indicates an increase of 1 point).