Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

The effect of chemotherapy on subjective cognitive function in younger early-stage breast cancer survivors treated with chemotherapy compared to older patients

The effect of chemotherapy on subjective cognitive function in younger early-stage breast cancer... Purpose To evaluate the impact of chemotherapy on subjective cognitive functioning according to age in a large cohort of breast cancer patients. Methods Within the UMBRELLA cohort, 715 patients with early-stage primary invasive breast cancer (T1-3N0-1M0) were selected. Subjective cognitive function was assessed by means of the EORTC QLQ-C30 up to 24 months and compared between patients treated with and without chemotherapy, for three different age strata (355 patients < 55 years, 240 patients aged 55–65 years, and 120 patients > 65 years). Differences between chemotherapy and non-chemotherapy patients by age at different time points were assessed by linear mixed-effect models correcting for age, tumor stage, educational level, endo- crine therapy, anxiety, and depression. Results In total, 979 patients from the UMBRELLA cohort were included, of which 715 (73%) responded to baseline and at least one follow-up questionnaire. Questionnaire response rates ranged between 92 and 70%. The proportion of patients treated with chemotherapy decreased with age: 64% (n = 277) in patients < 55 years, 45% (n = 107) in patients 55–65 years, and 23% (n = 27) in patients > 65 years. Chemotherapy was associated with reduced subjective cognitive functioning. The impact of chemotherapy on subjective cognitive function was most pronounced in patients < 55 years, followed by those between 55 and 65 years. In the youngest age groups, patients treated with chemotherapy had significantly lower cognitive functioning up to 24 months. In women over 65 years, subjective cognitive functioning was comparable between patients treated with and without chemotherapy. Conclusion This study confirms that chemotherapy is associated with impaired subjective self-reported cognitive function- ing in breast cancer patients, and the effect persists at least up to 2 years after diagnosis. The impact of chemotherapy on self-reported cognitive functioning in the first 24 months is most pronounced in younger patients, especially those under 55 years of age. Keywords Breast cancer · Chemotherapy · Cognitive function · Patient-reported outcome Introduction In the past decades, advances in early tumor detection, Preliminary data were presented at the Breast Cancer Meeting improved surgery, and more effective adjuvant treatment (Bossche Mammadagen) 2018 in The Netherlands. have led to a decreased mortality from breast cancer [1]. As Electronic supplementary material The online version of this the number of breast cancer survivors is consistently increas- article (https ://doi.org/10.1007/s1054 9-019-05149 -4) contains ing, long-term toxicity and morbidity following treatment supplementary material, which is available to authorized users. are a growing concern among patients and physicians [2, 3]. Adjuvant chemotherapy reduces breast cancer-related * M. L. Gregorowitsch m.l.gregorowitsch@umcutrecht.nl mortality [6]. However, chemotherapy is also the prime suspect to cause cancer-related cognitive decline [4–6]. Extended author information available on the last page of the article Vol.:(0123456789) 1 3 430 Breast Cancer Research and Treatment (2019) 175:429–441 This condition encompasses a range of symptoms including they had been diagnosed with a ductal carcinoma in situ, memory loss, inability to concentrate, difficulty in thinking, high-risk disease (cT4N2-3) or with clinical metastasis and other subtle cognitive changes [8]. Incidence rates of (M1). cognitive symptoms vary, but it has been suggested that up All patients were treated according to the Dutch guide- to 71% of breast cancer patients suffer from some degree lines for breast cancer treatment [13]. When chemotherapy of cognitive impairment after chemotherapy [3, 7, 8]. For was indicated, patients were treated with (neo)adjuvant some patients, symptoms of reduced cognitive function anthracycline-based chemotherapy and/or taxane-based may be transient, whereas for others it may persist and seri- chemotherapy. In case of Her2Neu receptor positivity, ously impact quality of life [9, 10]. Deterioration in cogni- trastuzumab (± pertuzumab) was given. All participants tive function is of particular importance in younger breast had undergone mastectomy or breast-conserving surgery cancer patients, given their longer life expectancy and the combined with axillary staging and/or surgery. All patients detrimental effect of cognition on employment opportunities were scheduled for radiotherapy of the breast or chest wall, and daily activities (e.g., taking care of children) [9, 10]. including regional radiotherapy if indicated. Adjuvant endo- Estimating chemotherapy-induced risks of cognitive dys- crine therapy was administered when indicated in hormone function is important to guide clinical and shared decision receptor positive patients. making, to adequately inform patients, and to offer targeted rehabilitation programs to those most at risk of developing cognitive problems. Numerous studies have aimed to analyze Data collection impairment in cancer survivors using hard endpoints, but only few studies have compared patient-reported cognitive Patient, tumor, and treatment characteristics were provided function in breast cancer patients and normative population by the Netherlands Cancer Registry (NCR) of the Nether- [11]. lands Comprehensive Cancer Organization (IKNL) [14]. The aim of the present study was to evaluate self-reported Data on subjective cognitive functioning were assessed subjective cognitive function in early-stage breast cancer using the cancer quality-of-life core questionnaire of the patients in different age groups up to 2 years after treatment, European Organization for Research and Treatment of and to compare outcomes between patients treated with and Cancer (EORTC QLQ-C30) [15]. The EORTC QLQ-C30 without chemotherapy. subscale for cognitive function includes two, 4-point Likert- scored items: (1) “Have you had difficulty in concentrating on things, like reading a newspaper or watching television?”; Methods and (2) “Have you had difficulty remembering things?” [EORTC Cognitive Functioning (EORTC-CF)]. Anxiety This study conducted with data from the Dutch prospective and depression were assessed using the Hospital Anxiety observational breast cancer cohort UMBRELLA (Utrecht and Depression Scale (HADS) [16]. Patient with scores of cohort for Multiple BREast cancer intervention studies and ≥ 8 were considered to have moderate of high probability of Long-term evaLuAtion) [12]. UMBRELLA was approved having anxiety or depressive disorders [17–19]. by the Medical Ethics Committee of the University Medi- The EORTC QLQ-C30 questionnaire was administered cal Center (UMC), Utrecht, the Netherlands, and is reg- at cohort entry (i.e., before radiotherapy, baseline) and at 3, istered on clinicaltrials.gov (NCT02839863). The cohort 6, 12, 18, and 24 months thereafter. The HADS was admin- includes patients (≥ 18 years) with histologically proven istered at all aforementioned time points except at 3 months. ductal carcinoma in situ or invasive breast cancer, who Due to enrollment at the Radiation Oncology Department, were referred to the Department of Radiation Oncology a minority of the patients undergoing neoadjuvant chemo- of the UMC, Utrecht, the Netherlands. All participants therapy already received treatment before cohort enrollment provided informed consent for the use of their clini- (i.e., baseline questionnaire). Questionnaires were collected cal data and the collection of patient-reported outcomes within the Patient-Reported Outcomes Following Initial (PRO) via standardized questionnaires. Informed consent treatment and Long-term Evaluation of Survivorship regis- was obtained after breast cancer surgery, before the start try (PROFILES) [20]. Subjective cognitive function scores of radiotherapy, or in a minority of the patients during of patients were compared with cross-sectional scores of an the course of neoadjuvant systemic treatment. For this age-matched Dutch reference population including women study, we identified all adult patients ≤ 70 years of age, without a history of breast cancer (total n = 944, Supplement enrolled between October 2013 and September 2017 and Table 1) provided by PROFILES [20]. Data of the reference who responded to at least two questionnaires assessing population were matched on age for the three different age PROs including baseline questionnaire (i.e., responders, strata (< 55 years n = 494, 55–65 years n = 247, > 65 years response rates: 70–92%). Patients were excluded when n = 203). 1 3 Breast Cancer Research and Treatment (2019) 175:429–441 431 and non-chemotherapy patients correcting for age, endocrine Statistical analysis treatment, educational level, anxiety and depression, and tumor stage, handling the correlation structures of repeated Patients were categorized according to chemotherapy sta- tus (i.e., non-chemotherapy and chemotherapy group) to measures nested within patients. In sensitivity analysis, we evaluated the effect of hormonal changes on subjective self- observe overall differences in subjective cognitive function. For this purpose, completion dates for questionnaires were reported cognitive functioning. We included pre/perimeno- pausal and postmenopausal status at baseline instead of age compared with the dates of the start of (neo)adjuvant chemo- therapy. Patients were categorized into the chemotherapy to evaluate if the effect of chemotherapy would change, as age is related to subjective cognitive impairment. A sec- group when the start of chemotherapy preceded the date of completing the questionnaire. Therefore, patients were ond sensitivity analysis was performed excluding patients treated with neoadjuvant chemotherapy (i.e., only including only categorized as chemotherapy patient, when chemo- therapy had been started. For example, patients receiving adjuvant chemotherapy patients). We included only patients treated with adjuvant chemotherapy to exclude the effect adjuvant chemotherapy 4 months after baseline were coded as ‘non-chemotherapy group’ at baseline and 3 months, but of more advanced tumor stage and more intense treatment regime, which are most likely patients receiving neoadjuvant as ‘chemotherapy group’ at 6, 12, 18, and 24 months. After 6 months, there were no more shifts from the non-chemo- chemotherapy, on cognitive function. As a measure of clinically meaningful difference, the therapy to the chemotherapy group. All patients that had started chemotherapy, remained in the chemotherapy group. standardized effect size (ES) was calculated (MD divided by the pooled standard deviation of the differences in scores) Next, patients were stratified by age in three groups, patients below 55 years, between 55 and 65 years, and above 65 years and classified as “no effect” (ES 0.2), “small effect” (ES, 0.2–0.4), “medium effect” (ES, 0.5–0.7), and “large effect” of age. Frequencies, proportions, and means with standard deviations for normally distributed variables (and medians (ES ≥ 0.8), according to Cohen [23]. A decline of 10 points on the EORTC QLQ-C30 cognitive function scale was con- with interquartile ranges (IQR) otherwise), were used to describe patient, tumor, and treatment characteristics. sidered clinically meaningful [24, 25]. The level of statistical significance was P < 0.05 and cor- The EORTC QLQ-C30 questionnaire was processed according to the EORTC scoring manual [21]. Cognitive rections for multiple testing were made. Statistical analyses were performed using SPSS Statistics for Windows, version function scores of the EORTC QLQ-C30 range from 0 to 100 with higher scores representing better cognitive func- 23 (IBM Corp, Armonk, NY). tioning [21]. Changes in cognitive function within the non- chemotherapy and chemotherapy within different age strata Results were analyzed with linear mixed-effects models to account for the correlation within subjects between the repeated In total, 1441 patients were enrolled in the cohort between measurements [22]. Linear mixed-effects models tested the presence of an October 2013 and September 2017 (Fig. 1). Of all partici- pants, 979 patients met the inclusion criteria and 715 (73%) age group-by-time-by-treatment (chemotherapy treatment versus no chemotherapy treatment) interactions for subjec- completed 2 or more quality-of-life (QOL) questionnaires. In total, 361 (50%) of the selected patients were treated with tive self-reported cognitive function. The model included a participant-specific random effect. Age, tumor stage, edu- either neoadjuvant (n = 131, 35%) or adjuvant (n = 230, 65%) chemotherapy. cational level, endocrine treatment, anxiety (HADS Anxi- ety score < 8 vs. ≥ 8), and depression (HADS Depression Patients in the chemotherapy group had a median age of 51 (IQR 13) and patients in the non-chemotherapy group had score < 8 vs. ≥ 8) were included as fixed effects to adjust for potential confounding effects. An autoregressive covariance a median age of 60 (IQR 14, Table 1). Patients treated with chemotherapy more often presented with a higher patho- structure of the first order (AR1) was used to define the cor - relations among observations, assuming correlations would logical tumor stage (T2 or T3) and clinical positive lymph nodes and had undergone more extensive treatment (e.g., be greater between measurements that were closer together in time compared with those further apart (i.e., exponential mastectomy, axillary lymph node dissection, and endocrine therapy). decline) [2]. Changes in subjective self-reported cognitive function are presented as the mean differences (MDs) with In total, half of the patients (n = 355) were aged 55 years or younger, 240 patients (33%) between 55 and 65 years and confidence intervals (95% CI), reflecting the difference between patients with chemotherapy and without chemo- 120 patients (17%) were older than 65 years. The majority of patients (n = 227, 64%) in the youngest age group received therapy treatment in the same age category group. The use of advanced modeling approaches allowed us to examine main chemotherapy, whereas 45% (n = 107) of the patients aged between 55 and 65 years and 23% (n = 27) of the > 65 years effects and interactions between age, time, chemotherapy, 1 3 432 Breast Cancer Research and Treatment (2019) 175:429–441 Fig. 1 Flowchart of patient inclusion within the UMBRELLA breast cancer cohort and response rates. Response rates were calcu- lated based on the opportunity patients had to return the questionnaire according to their inclusion date of age, were treated with chemotherapy (Table 2). Educa- reference population (n = 494), cognitive function of both tional level was highest in patients aged < 55 years of age. the chemotherapy and non-chemotherapy patients was lower at all time points. Comparison by treatment group In patients aged between 55 and 65 years, adjusted scores of subjective self-reported cognitive function were Subjective self-reported cognitive functioning, adjusted for significantly lower in those treated with chemotherapy at educational level, age, tumor stage, endocrine therapy, anxi- 3, 12, 18, and 24 months. Mean scores for subjective self- ety, and depression (HADS) of patients treated with chemo- reported cognitive function were significantly worse at all therapy, was significantly worse 3, 6, and 12 months after time intervals except for baseline scores and 6 months (MD the start of radiotherapy compared to non-chemotherapy 6.8, 95% CI 2.2–11.2, MD 8.1, 95% CI 3.1–13.0, MD 7.0, patients. In both groups, subjective self-reported cognitive 95% CI 1.6–12.3, MD 8.0, 95% CI 2.0–13.9, at 3, 12, 18, function deteriorated at 3 months when compared to base- and 24 months, resp., Table 3). The subjective self-reported line scores and improved thereafter with scores comparable cognitive score of the age-matched Dutch reference popula- to baseline scores at 24 months (Fig. 2). tion was higher. Among patients above 65 years of age, the difference in Comparison by age and treatment group cognitive function scores between chemotherapy groups was less pronounced (Fig. 3). At 3 months, patients treated Younger patients (< 55 years) treated with chemotherapy with chemotherapy had lower score compared to non- reported significantly worse subjective self-reported cogni- chemotherapy patients (MD 7.7, 95% CI − 4.3 to 19.7, ES; tive functioning than patients from the same age who did 0.3, Table  3). At 6, 12, 18, and 24 months, self-reported not receive chemotherapy treatment (Fig. 3), after adjusting subjective cognitive function scores were lower, but non- for age, tumor stage, educational level, endocrine therapy, significant, in patients not treated with chemotherapy. anxiety, and depression (HADS) in mixed model analysis The age-matched (> 65 years) Dutch reference population (Table 3). Mean differences (MD) in subjective self-reported showed comparable scores in chemotherapy patients at 18 cognitive functioning scores in patients < 55 years of age, and 24 months. treated with or without chemotherapy were largest at 6, 12, Sensitivity analysis of chemotherapy on subjective self- and 18 months (MD 12.3, 95% CI 8.2–16.3, MD 10.7, 95% reported cognitive function in pre/peri versus postmeno- CI 6.5–14.9, and MD 10.7, 95% CI 5.9–15.4, resp.). Effect pausal patients showed a similar effect of chemotherapy as sizes varied from 0.2 to 0.6, indicating a small/medium in the effect of chemotherapy in patients < 55 years of age effect (Table  2). Compared to the age-matched Dutch and patients between 55 and 65 years of age. The effect of 1 3 Breast Cancer Research and Treatment (2019) 175:429–441 433 Table 1 Demographics and Chemotherapy No chemotherapy disease characteristics of No. of patients (%) No. of patients (%) patients treated with and without chemotherapy in the Total no. of patients 361 (50) 354 (50) first 24 months after enrollment Age in years at inclusion, median (IQR) 51 (13) 60 (14) in the UMBRELLA breast cancer cohort Age category  < 55 years 227 (63) 128 (36)  55–65 years 107 (30) 133 (38)  > 65 years 27 (8) 93 (26) Menopausal status at inclusion  Premenopausal 158 (44) 58 (16)  Peri- or postmenopausal 183 (51) 269 (76)  Unknown 20 (6) 27 (8) Pathological tumor stage  T0 47 (13) 0 (0)  T1 181 (50) 316 (89)  T2 102 (28) 34 (10)  T3 20 (6) 0 (0)  Unknown 11 (3) 4 (1) Course of chemotherapy treatment  Neoadjuvant chemotherapy 135 (37) NA  Adjuvant chemotherapy 226 (63) NA Surgical treatment  Breast-conserving surgery 278 (77) 344 (97)  Mastectomy 83 (23) 10 (3) Most invasive axillary procedure  Sentinel lymph node biopsy 284 (79) 343 (97)  Axillary lymph node dissection 60 (17) 5 (1)  Unknown 17 (5) 6 (2) Estrogen receptor status  Negative 98 (27) 12 (3)  Positive 263 (73) 342 (97)  Unknown 0 (0) 2 (1) HER2 receptor status  Negative 270 (75) 338 (96)  Positive 91 (25) 6 (2)  Unknown 0 (0) 10 (3) Type of chemotherapy  Anthracycline based 328 (91) NA  Non-anthracycline based 34 (9) NA Endocrine treatment  No 109 (30) 218 (62)  Yes 252 (70) 135 (38) Type of radiotherapy  Local radiotherapy 224 (62) 324 (92)  Locoregional radiotherapy 137 (38) 29 (8)  Partial breast 0 (0) 1 (0) Highest educational le vel  Secondary school ± elementary school 79 (22) 96 (27)  Lower vocational education 82 (23) 68 (19)  Community college 111 (31) 84 (24)  University 61 (17) 50 (14)  Unknown 28 (8) 56 (16) 1 3 434 Breast Cancer Research and Treatment (2019) 175:429–441 Table 1 (continued) Chemotherapy No chemotherapy No. of patients (%) No. of patients (%) Working tasks  Physical work 35 (10) 34 (10)  Mental work 145 (40) 105 (30)  Combination of physical and mental work 97 (27) 73 (20)  Unknown 119 (33) 142 (40) HADS score on the Anxiety subscale, mean (SD) 6.2 (3.7) 5.5 (3.4) HADS score on the Depression subscale, mean (SD) 3.7 (3.5) 3.1 (3.3) All patients who received chemotherapy treatment in the first 24 months after enrollment in the UMBRELLA breast cancer cohort were considered chemotherapy patients and contributed to the chemo- therapy group Categories may not sum to total N or 100% because of missing values or rounding IQR interquartile range, NA not applicable, HER2 human epidermal growth factor receptor 2, HADS Hospi- tal Anxiety and Depression scale Neoadjuvant chemotherapy was given in combination with immunotherapy if patients were HER2 recep- tor positive Estrogen receptor positive > 10% Radiotherapy on the breast or chest wall with or without boost on the tumor bed Includes radiotherapy on axillary and/or periclavicular lymph nodes and/or internal mammary nodes Self-repor ted chemotherapy on subjective cognitive function was larg- to be more similar between patients who were exposed to est (ES: 0.4–0.6) in pre/perimenopausal patients (Supple- chemotherapy, not exposed to chemotherapy, and healthy ment Table 3). Sensitivity analysis on the effect of cognitive controls. functioning only including patients who received adjuvant Many cross-sectional and longitudinal studies have found chemotherapy patients shows bigger difference in cognitive evidence for chemotherapy-induced self-reported subjective function in the youngest patient group at baseline (MD 12.9, cognitive changes in patients with breast cancer; however, 95% CI; 3.7–22.0 Supplement Table 4). studies on the effect in different age categories are limited [11, 26–28]. In several studies on risk factors for cognitive decline, age has been found to be a well-established factor Discussion for subjective and objective cognitive deterioration. Even though most studies report that older patients may be more This prospective cohort study showed that self-reported sub- vulnerable to toxic effects of treatment, we did not see an jective cognitive function within the first 2 years after breast (strong) effect of chemotherapy on self-reported subjective cancer was lower in early-stage breast cancer patients treated cognitive function in patients aged over 65 years; however, with chemotherapy compared to patients treated without the number of older patients treated with chemotherapy chemotherapy with significantly lower scores at 3, 6, and was small in this study [2, 29–33]. Longitudinal study in 12 months. The short- and long-term ee ff ct of chemotherapy 112 breast cancer patients from Ahles et al. [32] found that on self-reported subjective cognitive function was most pro- older patients who were exposed to chemotherapy per- nounced in patients younger than 55 years of age. Although formed worse on neuropsychological measures of processing self-reported subjective cognitive functioning of younger speed compared to non-chemotherapy patients and healthy patients slightly improved 3  months after treatment, the controls, i.e., older patients had lower objective cognitive chemotherapy group consistently reported statistically sig- functioning. Lower pretreatment cognitive reserve may nificant worse scores. Up to 2 years after diagnosis, younger make older patients more prone to the negative influence of patients (≤ 65 years) experienced worse self-reported subjec- chemotherapy on brain function, resulting in small cogni- tive cognitive functioning compared to the Dutch reference tive changes in cognition being detectable more easily with population, matched on age reflecting the same age of the neuropsychological tests [34]. Although Ahles et al. [32] two youngest age strata (respectively, < 55 and 55–65 years). corrected for cognitive reserve at baseline, they did not find This was in contrast with older patients (> 65  years), in a significant relation with older age and self-reported subjec- whom self-reported subjective cognitive function was found tive cognitive function, solely chemotherapy treatment was 1 3 Breast Cancer Research and Treatment (2019) 175:429–441 435 Table 2 Demographics and disease characteristics of patients below 55 years of age, between 55 and 65 years of age, and above the age of 65 years, participating in the UMBRELLA breast cancer cohort < 55 years 55–65 years > 65 years No. of patients (%) No. of patients (%) No. of patients (%) Total no. of patients 355 (50) 240 (34) 120 (17) Total no. of patients treated with chemotherapy 227 (64) 107 (45) 27 (23) Age at inclusion, median (IQR) 49 (6) 60 (4) 68 (2)  Menopausal status at inclusion  Premenopausal 213 (60) 3 (1) 0 (0)  Peri- or postmenopausal 99 (28) 233 (97) 120 (100)  Unknown 43 (12) 4 (2) 0 (0) Pathological tumor stage  T0 32 (9) 12 (5) 3 (3)  T1 226 (64) 176 (73) 95 (79)  T2 71 (20) 45 (19) 20 (17)  T3 17 (5) 2 (1) 1 (1)  Unknown 9 (3) 5 (2) 1 (1) Course of chemotherapy treatment  Neoadjuvant chemotherapy 102 (29) 25 (10) 7 (6)  Adjuvant chemotherapy 125 (35) 82 (34) 20 (17) Surgical treatment  Breast-conserving surgery 291 (82) 220 (92) 111 (93)  Mastectomy 64 (18) 20 (8) 9 (8) Most invasive axillary procedure  Sentinel lymph node biopsy 302 (85) 216 (90) 109 (91)  Axillary lymph node dissection 41 (12) 17 (7) 7 (6)  Unknown 12 (3) 7 (3) 4 (3) Estrogen receptor status  Negative 61 (17) 32 (13) 17 (14)  Positive 294 (83) 207 (86) 102 (85)  Unknown 0 (0) 1 (0) 1 (1) HER2 receptor status  Negative 286 (81) 213 (89) 109 (91)  Positive 65 (18) 23 (10) 9 (8)  Unknown 4 (1) 2 (1) 1 (1) Adjuvant chemotherapy treatment  No 230 (65) 158 (66) 100 (83)  Yes 125 (35) 82 (34) 20 (17) Type of chemotherapy  Anthracycline based 207 (58) 99 (41) 23 (19)  Non-anthracycline based 21 (6) 8 (3) 5 (4) Endocrine treatment  No 144 (41) 115 (48) 69 (58)  Yes 211 (59) 125 (52) 51 (43) Type of radiotherapy  Local radiotherapy 254 (72) 194 (81) 100 (83)  Locoregional radiotherapy 101 (28) 45 (19) 20 (17)  Partial breast 0 (0) 1 (0) 0 (0) Highest educational level  Secondary school ± elementary school 66 (19) 77 (32) 32 (27)  Lower vocational education 93 (26) 48 (20) 9 (8)  Community college 109 (31) 73 (30) 13 (11) 1 3 436 Breast Cancer Research and Treatment (2019) 175:429–441 Table 2 (continued) < 55 years 55–65 years > 65 years No. of patients (%) No. of patients (%) No. of patients (%)  University 82 (23) 23 (10) 6 (5)  Unknown 5 (1) 19 (8) 60 (50) HADS score on the Anxiety subscale, mean (SD) 6.2 (3.7) 5.7 (3.7) 5.2 (3.4) HADS score on the Depression subscale, mean (SD) 3.6 (3.5) 3.3 (3.4) 3.1 (3.3) Categories may not sum to total N because of missing values NA not applicable, HER2 human epidermal growth factor receptor 2, HADS Hospital Anxiety and Depression scale Neoadjuvant chemotherapy was given in combination with immunotherapy if patients were HER2 receptor positive Estrogen receptor positive > 10% Radiotherapy on the breast or chest wall with or without boost on the tumor bed Includes radiotherapy on axillary and/or periclavicular lymph nodes and/or internal mammary nodes Self-repor ted Fig. 2 Cognitive functioning. Higher score indicates better cognitive functioning found to be significantly related to self-reported subjective Results showed that self-reported cognitive impairment is cognitive function. Mandelblatt et al. [29] recently showed a substantial and pervasive problem for patients during and that subjective self-reported cognition for survivors (mean after chemotherapy treatment [28]. Younger age and black age 68) exposed to chemotherapy decreases over time, race were associated with problems with perceived cognitive whereas cognition of non-exposed patients did not change; abilities [28]. As cancer treatment, like chemotherapy, may however, the decline among the chemo exposed was non- accelerate the aging process via processes like DNA dam- significant. They found that older patients with genotypes age, it is plausible that the greatest effect of chemotherapy associated with neurodegenerative disease, such as apoli- on cognitive function is detected in younger patients [36]. poprotein E, exposed to chemotherapy showed a clinically Furthermore, it might be that younger patients may detect meaningful decrease in self-reported cognitive function at impairment more readily as they are less likely to attribute 24 months [29]. These results support the idea that not only problems to age-related changes. Also, such subjective cog- aging but also aging phenotypes are associated with lower nitive problems are less likely to be obscured by preexist- self-reported cognitive function. It has been suggested that ing age-related deficits [31]. Younger women may perceive chemotherapy can lead to cancer-related cognitive declines subjective cognitive impairment due to treatment more often through acceleration of aging processes [35]. than older patients because they are more often employed Our study indicates that chemotherapy predominantly has and working, which requires more from their cognition and a substantial detrimental effect on subjective self-reported intellect. cognition in younger women (particularly < 55  years). Besides chemotherapy, other factors may also affect sub- Janelsins et al. compared subjective self-reported cognitive jective cognitive functioning in patients. Also, in women function of patients exposed to chemotherapy (n = 581), ≤ 65 years treated without chemotherapy, cognitive function- mean age of 53 years, to age-matched non-cancer controls. ing was lower compared to healthy age-matched controls. 1 3 Breast Cancer Research and Treatment (2019) 175:429–441 437 Fig. 3 Cognitive function in early-stage breast cancer patients according to chemo- therapy treatment stratified by age categories and compared to the age-matched Dutch reference population. Cogni- tive function was assessed with the cancer questionnaire of the European Organization for Research and Treatment of Cancer (EORTC QLQ- C30). Results are presented as adjusted mean scores accom- panied with 95% confidence intevals. Higher score indicates better cognitive function The impact of a cancer diagnosis itself, surgery, and tox- et al., reported that psychological consequences of a cancer icity due to other treatments may also have effect on the diagnosis and treatment contribute more to cognitive dys- reporting of cognitive symptoms [37]. Furthermore, sub- function than the effects of medication [37]. The feeling of jective cognitive changes are strongly associated with other life disruption and sick leave might be more pronounced in patient-reported outcomes like fatigue, anxiety, and depres- younger patients contributing to greater cognitive impair- sion making it challenging to determine whether cognitive ment compared to older patients [37]. functioning is solely related to underlying brain dysfunc- Sensitivity analysis was performed stratifying by meno- tion [11]. Longitudinal study in 187 breast cancer patients pausal status instead of age. Reason behind this is that evi- with and without chemotherapy exposure from Hermelink dence has emerged suggesting that hormonal changes, alone 1 3 438 Breast Cancer Research and Treatment (2019) 175:429–441 Table 3 Cognitive function in younger (< 55 years), middle-aged (55–65 years), and older (> 65 years) patients with breast cancer assessed with the EORTC QLQ-C30 at baseline, 3, 6, 12, 18, and 24 months after first consultation with the radiation oncologist Age < 55 years Age 55–65 years Age > 65 years Overall P value No chemotherapy n = 128 n = 133 n = 93 (reference group) Chemotherapy n = 277 n = 107 n = 27 a b c a b c a b c MD 95% CI P value ES MD 95% CI P value ES MD 95% CI P value ES Baseline 5.6 1.9–9.3 0.003 0.2 2.9 − 4.1 to 9.9 0.422 0.1 1.7 − 16.6 to 20.1 0.853 0.0 0.017 3 Months 4.4 0.5–8.2 0.024 0.2 6.8 2.2–11.2 0.003 0.3 7.7 − 4.3 to 19.7 0.208 0.3 6 Months 12.3 8.2–16.3 0.000 0.6 4.4 − 0.3 to 9.1 0.072 0.2 − 4.7 − 16.8 to 7.5 0.454 0.2 12 Months 10.7 6.5–14.9 0.000 0.5 8.1 3.1–13.0 0.001 0.4 − 1.8 − 14.2 to 10.5 0.772 0.1 18 Months 10.7 5.9–15.4 0.000 0.5 7.0 1.6–12.3 0.010 0.3 − 9.9 − 23.3 to 3.6 0.151 0.4 24 Months 7.4 1.8–13.2 0.009 0.4 8.0 2.0–13.9 0.009 0.4 − 9.5 − 25.1 to 6.1 0.235 0.3 The differences in mean score (MD) between in the younger, middle-aged, and older patient group are adjusted for endocrine treatment and show the difference in cognitive function between chemotherapy and non-chemotherapy treatment Patient-reported outcomes on cognitive functioning according to EORTC QLQ-C30 Between-group effects were assessed using mixed models including the measurements obtained at baseline and at 3, 6, 12, 18, and 24 months, adjusted for multiple testing, age, tumor stage, endocrine treatment, educational level, anxiety (HADS), and depression (HADS). Patients not exposed to chemotherapy serve as a reference category to calculate mean differences N number of patients, MD mean difference, 95% CI confidence interval, ES effect size Difference in mean score with reference group The P value shown for the difference between no chemotherapy and chemotherapy group Standardized effect size calculated (mean difference divided by the pooled standard deviation) as a measure for minimal clinically important change. Small effect if ES 0.2–0.4, medium 0.5–0.7, large effect > 0.8 The P value shown for the interaction Age-by-time by treatment (no chemotherapy versus chemotherapy) or in combination with chemotherapy and endocrine therapy, genomic testing to guide decisions on withholding chemo- may cause cognitive impairment [29, 38–40]. An observed therapy in selected patients. Furthermore, results of this effect could thus be an age-related effect rather than an inde- study, can help patients and doctors in shared decision mak- pendent effect of systemic treatment per se. Results showed ing about chemotherapy treatment. As many young patients that the ee ff ct of chemotherapy was largest in pre/perimeno - have to return to work, they should be adequately informed pausal women treated with chemotherapy. by physicians about the risks of cognitive decline during and This study adds to the growing literature suggesting after treatment with chemotherapy. that patients experience subjective self-reported cognitive A limitation of our study is the fact that the problems after cancer and/or cancer-related treatments like UMBRELLA cohort was not specifically designed to eval- chemotherapy. The effect of chemotherapy on subjective uate the effect of chemotherapy, and patients were enrolled cognitive function, especially in young patients at 6, 12, in our study at their first consultation with the radiation and 18 months after baseline, emphasizes the importance oncologist instead. In the majority of the patients, this was of adequate selection of early-stage breast cancer patients after breast surgery before the start of radiotherapy treat- whom benefit most from chemotherapy treatment and to ment. However, not all patients were enrolled and com- avoid overtreatment with chemotherapy and its attendant pleted the baseline PRO questionnaires, at the same time toxic effects. In recent years, research has been devoted to in their treatment trajectory (i.e., after surgery, before the the development and validation of genomic tests that can start of radiotherapy). This especially applies for women provide not only prognostic information but perhaps more who had received neoadjuvant chemotherapy treatment importantly can predict response to therapy [41]. Oncotype and were enrolled before starting radiotherapy treatment DX and MammaPrint are well validated and the most widely which were mostly younger patients. Although we tried to used multigene signatures for predicting outcomes in breast correct for these differences in treatment trajectories, we cancer [42, 43]. The use of genomic testing to guide chemo- might have missed early psychological complaints due to therapy treatment has been shown to lead to a reduction in cancer diagnosis and neoadjuvant chemotherapy in some the use of adjuvant chemotherapy in patients with early- patients. Furthermore, the lack of pretreatment subjective stage breast cancer [42]. This highlights the importance of cognitive function, which prevented us from adjusting 1 3 Breast Cancer Research and Treatment (2019) 175:429–441 439 Informed consent Informed consent was obtained from all individual for pretreatment cognitive function, may determine the participants included in the study. amount of cognitive decline. The use of the EORTC Cognitive Functioning alone, to assess subjective self- Open Access This article is distributed under the terms of the Crea- reported cognitive function, may result in an underesti- tive Commons Attribution 4.0 International License (http://creat iveco mation of the extent of an individual’s cognitive symptoms mmons.or g/licenses/b y/4.0/), which permits unrestricted use, distribu- as this EORTC domain encompasses and only captures tion, and reproduction in any medium, provided you give appropriate two aspects of cognition (concentration and memory) credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. [11]. Also, self-assessment of cognitive functioning may be prone to nocebo effects. In other words, information about the association between cognitive difficulties and chemotherapy or endocrine therapy given to chemotherapy References patients and patients who received endocrine therapy (± chemotherapy) might increase the reporting of cognitive 1. Holzel D, Eckel R, Bauerfeind I, Baier B, Beck T, Braun M problems [44]. No formal neuropsychological tests were et al (2017) Improved systemic treatment for early breast can- cer improves cure rates, modifies metastatic pattern and shortens performed to test objective cognitive functioning. Objec- post-metastatic survival: 35-year results from the Munich Cancer tive cognitive functioning might have provided us with Registry. J Cancer Res Clin Oncol 143(9):1701–1712 a different association between chemotherapy and the 2. Ganz PA, Kwan L, Castellon SA, Oppenheim A, Bower JE, Sil- effect on cognition since self-reported cognitive symptoms verman DH et al (2013) Cognitive complaints after breast cancer treatments: examining the relationship with neuropsychological are known to be strongly associated with other patient- test performance. J Natl Cancer Inst 105(11):791–801 reported outcomes like fatigue, anxiety, and depression 3. Boykoff N, Moieni M, Subramanian SK (2009) Confronting [11]. Although we corrected for the use of endocrine ther- chemobrain: an in-depth look at survivors’ reports of impact on apy, we do not have information on compliance to adjuvant work, social networks, and health care response. J Cancer Surviv 3(4):223–232 endocrine treatment. A substantial proportion of patients 4. Winocur G, Johnston I, Castel H (2018) Chemotherapy and cogni- may have not been compliant to endocrine treatment due tion: International Cognition and Cancer Task Force recommen- to side effects. dations for harmonising preclinical research. Cancer Treat Rev In conclusion, patients treated with chemotherapy 69:72–83 5. Early Breast Cancer Trialists’ Collaborative Group (EBCTCG), report more impaired subjective cognitive functioning Peto R, Davies C, Godwin J, Gray R, Pan HC et  al (2012) as compared to those treated without chemotherapy. This Comparisons between different polychemotherapy regimens effect starts shortly after treatment and is seen up until 24 for early breast cancer: meta-analyses of long-term outcome months of follow-up. The association between chemother- among 100,000 women in 123 randomised trials. Lancet 379(9814):432–444 apy and impaired subjective cognitive functioning is most 6. Wefel JS, Vardy J, Ahles T, Schagen SB (2011) International pronounced in younger women, while in older patients dif- Cognition and Cancer Task Force recommendations to harmonise ferences in subjective cognitive function between exposed studies of cognitive function in patients with cancer. Lancet Oncol and non-exposed to chemotherapy were less. These results 12(7):703–708 7. Wefel JS, Saleeba AK, Buzdar AU, Meyers CA (2010) Acute and highlight the careful consideration that is needed in the late onset cognitive dysfunction associated with chemotherapy in current clinical decision making whether chemotherapy women with breast cancer. Cancer 116(14):3348–3356 should be administered to patients with early-stage breast 8. Pullens MJ, De Vries J, Van Warmerdam LJ, Van De Wal MA, cancer. Furthermore, with the emergence of increasing Roukema JA (2013) Chemotherapy and cognitive complaints in women with breast cancer. Psychooncology 22(8):1783–1789 evidence for the efficacy of cognitive training strategies, 9. Schagen SB, Muller MJ, Boogerd W, Rosenbrand RM, van Rhijn there should be more management options available to D, Rodenhuis S et al (2002) Late effects of adjuvant chemotherapy cancer patients to address this important issue. on cognitive function: a follow-up study in breast cancer patients. Ann Oncol 13(9):1387–1397 Acknowledgements We would like to thank Rebecca Stellato for her 10. Koppelmans V, Breteler MM, Boogerd W, Seynaeve C, Gundy C, statistical advice on mixed model analysis. Schagen SB (2012) Neuropsychological performance in survivors of breast cancer more than 20 years after adjuvant chemotherapy. J Clin Oncol 30(10):1080–1086 Compliance with ethical standards 11. Bray VJ, Dhillon HM, Vardy JL (2018) Systematic review of self- reported cognitive function in cancer patients following chemo- Conflict of interest The authors declare that they have no conflict of therapy treatment. J Cancer Surviv 12(4):537–559 interest. 12. Young-Afat DA, van Gils CH, van den Bongard HJGD, Verkooi- jen HM, UMBRELLA Study Group (2017) The Utrecht cohort Ethical approval All procedures performed in studies involving human for Multiple BREast cancer intervention studies and Long-term participants were in accordance with the ethical standards of the insti- evaLuAtion (UMBRELLA): objectives, design, and baseline tutional and/or national research committee and with the 1964 Helsinki results. Breast Cancer Res Treat 164(2):445–450 declaration and its later amendments or comparable ethical standards. 1 3 440 Breast Cancer Research and Treatment (2019) 175:429–441 13. Risicoprofilering Richtlijn Mammacarcinoom (2012) Kwaliteit- of breast cancer: CALGB 369901 (Alliance). Cancer 2016 sinstituut voor de gezondheidszorg CBO VvIK, pp 81–83 122(22):3555–3563 14. Netherlands Cancer Registry (2018) http://www.cijfe rsove rkank 31. Debess J, Riis JO, Engebjerg MC, Ewertz M (2010) Cogni- er.nl. Accessed May 2018 tive function after adjuvant treatment for early breast cancer: a 15. Aaronson NK, Ahmedzai S, Bergman B, Bullinger M, Cull A, population-based longitudinal study. Breast Cancer Res Treat Duez NJ et al (1993) The European Organization for Research 121(1):91–100 and Treatment of Cancer QLQ-C30: a quality-of-life instrument 32. Ahles TA, Saykin AJ, McDonald BC, Li Y, Furstenberg CT, for use in international clinical trials in oncology. J Natl Cancer Hanscom BS et  al (2010) Longitudinal assessment of cogni- Inst 85(5):365–376 tive changes associated with adjuvant treatment for breast 16. Spinhoven P, Ormel J, Sloekers PP, Kempen GI, Speckens AE, cancer: impact of age and cognitive reserve. J Clin Oncol Van Hemert AM (1997) A validation study of the Hospital Anxi- 28(29):4434–4440 ety and Depression Scale (HADS) in different groups of Dutch 33. Hurria A, Hurria A, Zuckerman E, Panageas KS, Fornier M, subjects. Psychol Med;27(2):363–370 D’Andrea G et al (2006) A prospective, longitudinal study of the 17. Zigmond AS, Snaith RP (1983) The hospital anxiety and depres- functional status and quality of life of older patients with breast sion scale. Acta Psychiatr Scand 67(6):361–370 cancer receiving adjuvant chemotherapy. J Am Geriatr Soc 18. Smith AB, Selby PJ, Velikova G, Stark D, Wright EP, Gould A 54(7):1119–1124 et al (2002) Factor analysis of the Hospital Anxiety and Depres- 34. Healey MK, Campbell KL, Hasher L (2008) Cognitive aging and sion Scale from a large cancer population. Psychol Psychother increased distractibility: costs and potential benefits. Prog Brain 75(Pt 2):165–176 Res 169:353–363 19. Snaith RP (2003) The Hospital Anxiety And Depression Scale. 35. Vega JN, Dumas J, Newhouse PA (2018) Self-reported chemother- Health Qual Life Outcomes 1:29–7525 apy-related cognitive impairment compared with cognitive com- 20. van de Poll-Franse LV, Horevoorts N, van Eenbergen M, Denollet plaints following menopause. Psychooncology 27(9):2198–2205 J, Roukema JA, Aaronson NK et al (2011) The Patient Reported 36. Ahles TA, Root JC, Ryan EL (2012) Cancer- and cancer treat- Outcomes Following Initial treatment and Long term Evaluation ment-associated cognitive change: an update on the state of the of Survivorship registry: scope, rationale and design of an infra- science. J Clin Oncol 30(30):3675–3686 structure for the study of physical and psychosocial outcomes in 37. Hermelink K, Buhner M, Sckopke P, Neufeld F, Kaste J, Voigt V cancer survivorship cohorts. Eur J Cancer 47(14):2188–2194 et al (2017) Chemotherapy and post-traumatic stress in the cau- 21. Fayers PM, Aaronson NK (2001) The EORTC QLQ-C30 scoring sation of cognitive dysfunction in breast cancer patients. J Natl manual, 3rd edn. European Organisation for Research and Treat- Cancer Inst. https ://doi.org/10.1093/jnci/djx05 7 ment of Cancer, Brussels 38. Mandelblatt JS, Jacobsen PB, Ahles T (2014) Cognitive effects of 22. Bonnetain F, Fiteni F, Efficace F, Anota A (2016)Statistical chal- cancer systemic therapy: implications for the care of older patients lenges in the analysis of health-related quality of life in cancer and survivors. J Clin Oncol 32(24):2617–2626 clinical trials. J Clin Oncol 34(16):1953–1956 39. Schilder CM, Eggens PC, Seynaeve C, Linn SC, Boogerd W, 23. ’Cohen J (1988) Statistical power analysis for the behavioral sci- Gundy CM et  al (2009) Neuropsychological functioning in ences, 2nd edn. Lawrence Erlbaum, Hillsdale postmenopausal breast cancer patients treated with tamoxifen 24. Bedard G, Zeng L, Lam H, Cella D, Zhang L, Lauzon N et al or exemestane after AC-chemotherapy: cross-sectional findings (2012) Meaningful change in oncology quality-of-life instru- from the neuropsychological TEAM-side study. Acta Oncol ments: a systematic literature review. Expert Rev Pharmacoecon 48(1):76–85 Outcomes Res 12(4):475–483 40. Mandelblatt JS, Clapp JD, Luta G, Faul LA, Tallarico MD, 25. Raman S, Ding K, Chow E, Meyer RM, van der Linden YM, McClendon TD et  al (2016) Long-term trajectories of self- Roos D et al (2018) Minimal clinically important differences in reported cognitive function in a cohort of older survivors of breast the EORTC QLQ-C30 and brief pain inventory in patients under- cancer: CALGB 369901 (Alliance). Cancer 122(22):3555–3563 going re-irradiation for painful bone metastases. Qual Life Res 22 27(4):1089–1098 41. Nicolini A, Ferrari P, Duffy MJ (2018) Prognostic and predic- 26. Jung MS, Zhang M, Askren MK, Berman MG, Peltier S, Hayes tive biomarkers in breast cancer: past, present and future. Semin DF et al (2017) Cognitive dysfunction and symptom burden in Cancer Biol 52(Pt 1):56–73 women treated for breast cancer: a prospective behavioral and 42. Cardoso F, van’t Veer LJ, Bogaerts J, Slaets L, Viale G, Delaloge fMRI analysis. Brain Imaging Behav 11(1):86–97 S et al (2016) 70-Gene signature as an aid to treatment decisions 27. Quesnel C, Savard J, Ivers H (2009) Cognitive impairments asso- in early-stage breast cancer. N Engl J Med 375(8):717–729 ciated with breast cancer treatments: results from a longitudinal 43. Sparano JA, Gray RJ, Makower DF, Pritchard KI, Albain KS, study. Breast Cancer Res Treat 116(1):113–123 Hayes DF et al (2015) Prospective validation of a 21-gene expres- 28. Janelsins MC, Heckler CE, Peppone LJ, Kamen C, Mustian KM, sion assay in breast cancer. N Engl J Med 373(21):2005–2014 Mohile SG et  al (2017) Cognitive complaints in survivors of 44. Jacobs W, Das E, Schagen SB (2017) Increased cognitive problem breast cancer after chemotherapy compared with age-matched reporting after information about chemotherapy-induced cognitive controls: an analysis from a nationwide, multicenter, prospective decline: The moderating role of stigma consciousness. Psychol longitudinal study. J Clin Oncol 35(5):506–514 Health 32(1):78–93 29. Mandelblatt JS, Small BJ, Luta G, Hurria A, Jim H, McDonald BC et al (2018) Cancer-related cognitive outcomes among older Publisher’s Note Springer Nature remains neutral with regard to breast cancer survivors in the thinking and living with cancer jurisdictional claims in published maps and institutional affiliations. study. J Clin Oncol. https ://doi.org/10.1200/JCO.18.00140 30. Mandelblatt JS, Clapp JD, Luta G, Faul LA, Tallarico MD, McClendon TD et  al (2016) Long-term trajectories of self- reported cognitive function in a cohort of older survivors 1 3 Breast Cancer Research and Treatment (2019) 175:429–441 441 Affiliations 1 2 1,3 4 4 5 M. L. Gregorowitsch  · A. Ghedri  · D. A. Young‑Afat  · R. Bijlsma  · I. O. Baas  · C. van Schaikv ‑ an de Mheen  · 6 7 8 1 2,9 M. J. Agterof  · E. Göker  · D. ten Bokkel Huinink  · H. J. G. D. van den Bongard  · H. M. Verkooijen  · UMBRELLA study group 1 6 Department of Radiation Oncology, University Department of Medical Oncology, St. Antonius, Nieuwegein, Medical Center (UMC) Utrecht, Heidelberglaan 100, The Netherlands 3584 CX Utrecht, The Netherlands Department of Medical Oncology, Alexander Monro Clinics, University of Utrecht, Utrecht, The Netherlands Bilthoven, The Netherlands 3 8 Department of Epidemiology, Julius Center for Health Department of Medical Oncology, Diakonessenhuis, Utrecht, Sciences and Primary Care, University Medical Center, The Netherlands Utrecht, The Netherlands Imaging Division, University Medical Center Utrecht, Department of Medical Oncology, University Medical Center Utrecht, The Netherlands Utrecht, Utrecht, The Netherlands Department of Medical Oncology, Meander Medical Center, Amersfoort, The Netherlands 1 3 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Breast Cancer Research and Treatment Springer Journals

The effect of chemotherapy on subjective cognitive function in younger early-stage breast cancer survivors treated with chemotherapy compared to older patients

Loading next page...
 
/lp/springer-journals/the-effect-of-chemotherapy-on-subjective-cognitive-function-in-younger-XmiYKZ0wqb

References (48)

Publisher
Springer Journals
Copyright
Copyright © 2019 by The Author(s)
Subject
Medicine & Public Health; Oncology
ISSN
0167-6806
eISSN
1573-7217
DOI
10.1007/s10549-019-05149-4
Publisher site
See Article on Publisher Site

Abstract

Purpose To evaluate the impact of chemotherapy on subjective cognitive functioning according to age in a large cohort of breast cancer patients. Methods Within the UMBRELLA cohort, 715 patients with early-stage primary invasive breast cancer (T1-3N0-1M0) were selected. Subjective cognitive function was assessed by means of the EORTC QLQ-C30 up to 24 months and compared between patients treated with and without chemotherapy, for three different age strata (355 patients < 55 years, 240 patients aged 55–65 years, and 120 patients > 65 years). Differences between chemotherapy and non-chemotherapy patients by age at different time points were assessed by linear mixed-effect models correcting for age, tumor stage, educational level, endo- crine therapy, anxiety, and depression. Results In total, 979 patients from the UMBRELLA cohort were included, of which 715 (73%) responded to baseline and at least one follow-up questionnaire. Questionnaire response rates ranged between 92 and 70%. The proportion of patients treated with chemotherapy decreased with age: 64% (n = 277) in patients < 55 years, 45% (n = 107) in patients 55–65 years, and 23% (n = 27) in patients > 65 years. Chemotherapy was associated with reduced subjective cognitive functioning. The impact of chemotherapy on subjective cognitive function was most pronounced in patients < 55 years, followed by those between 55 and 65 years. In the youngest age groups, patients treated with chemotherapy had significantly lower cognitive functioning up to 24 months. In women over 65 years, subjective cognitive functioning was comparable between patients treated with and without chemotherapy. Conclusion This study confirms that chemotherapy is associated with impaired subjective self-reported cognitive function- ing in breast cancer patients, and the effect persists at least up to 2 years after diagnosis. The impact of chemotherapy on self-reported cognitive functioning in the first 24 months is most pronounced in younger patients, especially those under 55 years of age. Keywords Breast cancer · Chemotherapy · Cognitive function · Patient-reported outcome Introduction In the past decades, advances in early tumor detection, Preliminary data were presented at the Breast Cancer Meeting improved surgery, and more effective adjuvant treatment (Bossche Mammadagen) 2018 in The Netherlands. have led to a decreased mortality from breast cancer [1]. As Electronic supplementary material The online version of this the number of breast cancer survivors is consistently increas- article (https ://doi.org/10.1007/s1054 9-019-05149 -4) contains ing, long-term toxicity and morbidity following treatment supplementary material, which is available to authorized users. are a growing concern among patients and physicians [2, 3]. Adjuvant chemotherapy reduces breast cancer-related * M. L. Gregorowitsch m.l.gregorowitsch@umcutrecht.nl mortality [6]. However, chemotherapy is also the prime suspect to cause cancer-related cognitive decline [4–6]. Extended author information available on the last page of the article Vol.:(0123456789) 1 3 430 Breast Cancer Research and Treatment (2019) 175:429–441 This condition encompasses a range of symptoms including they had been diagnosed with a ductal carcinoma in situ, memory loss, inability to concentrate, difficulty in thinking, high-risk disease (cT4N2-3) or with clinical metastasis and other subtle cognitive changes [8]. Incidence rates of (M1). cognitive symptoms vary, but it has been suggested that up All patients were treated according to the Dutch guide- to 71% of breast cancer patients suffer from some degree lines for breast cancer treatment [13]. When chemotherapy of cognitive impairment after chemotherapy [3, 7, 8]. For was indicated, patients were treated with (neo)adjuvant some patients, symptoms of reduced cognitive function anthracycline-based chemotherapy and/or taxane-based may be transient, whereas for others it may persist and seri- chemotherapy. In case of Her2Neu receptor positivity, ously impact quality of life [9, 10]. Deterioration in cogni- trastuzumab (± pertuzumab) was given. All participants tive function is of particular importance in younger breast had undergone mastectomy or breast-conserving surgery cancer patients, given their longer life expectancy and the combined with axillary staging and/or surgery. All patients detrimental effect of cognition on employment opportunities were scheduled for radiotherapy of the breast or chest wall, and daily activities (e.g., taking care of children) [9, 10]. including regional radiotherapy if indicated. Adjuvant endo- Estimating chemotherapy-induced risks of cognitive dys- crine therapy was administered when indicated in hormone function is important to guide clinical and shared decision receptor positive patients. making, to adequately inform patients, and to offer targeted rehabilitation programs to those most at risk of developing cognitive problems. Numerous studies have aimed to analyze Data collection impairment in cancer survivors using hard endpoints, but only few studies have compared patient-reported cognitive Patient, tumor, and treatment characteristics were provided function in breast cancer patients and normative population by the Netherlands Cancer Registry (NCR) of the Nether- [11]. lands Comprehensive Cancer Organization (IKNL) [14]. The aim of the present study was to evaluate self-reported Data on subjective cognitive functioning were assessed subjective cognitive function in early-stage breast cancer using the cancer quality-of-life core questionnaire of the patients in different age groups up to 2 years after treatment, European Organization for Research and Treatment of and to compare outcomes between patients treated with and Cancer (EORTC QLQ-C30) [15]. The EORTC QLQ-C30 without chemotherapy. subscale for cognitive function includes two, 4-point Likert- scored items: (1) “Have you had difficulty in concentrating on things, like reading a newspaper or watching television?”; Methods and (2) “Have you had difficulty remembering things?” [EORTC Cognitive Functioning (EORTC-CF)]. Anxiety This study conducted with data from the Dutch prospective and depression were assessed using the Hospital Anxiety observational breast cancer cohort UMBRELLA (Utrecht and Depression Scale (HADS) [16]. Patient with scores of cohort for Multiple BREast cancer intervention studies and ≥ 8 were considered to have moderate of high probability of Long-term evaLuAtion) [12]. UMBRELLA was approved having anxiety or depressive disorders [17–19]. by the Medical Ethics Committee of the University Medi- The EORTC QLQ-C30 questionnaire was administered cal Center (UMC), Utrecht, the Netherlands, and is reg- at cohort entry (i.e., before radiotherapy, baseline) and at 3, istered on clinicaltrials.gov (NCT02839863). The cohort 6, 12, 18, and 24 months thereafter. The HADS was admin- includes patients (≥ 18 years) with histologically proven istered at all aforementioned time points except at 3 months. ductal carcinoma in situ or invasive breast cancer, who Due to enrollment at the Radiation Oncology Department, were referred to the Department of Radiation Oncology a minority of the patients undergoing neoadjuvant chemo- of the UMC, Utrecht, the Netherlands. All participants therapy already received treatment before cohort enrollment provided informed consent for the use of their clini- (i.e., baseline questionnaire). Questionnaires were collected cal data and the collection of patient-reported outcomes within the Patient-Reported Outcomes Following Initial (PRO) via standardized questionnaires. Informed consent treatment and Long-term Evaluation of Survivorship regis- was obtained after breast cancer surgery, before the start try (PROFILES) [20]. Subjective cognitive function scores of radiotherapy, or in a minority of the patients during of patients were compared with cross-sectional scores of an the course of neoadjuvant systemic treatment. For this age-matched Dutch reference population including women study, we identified all adult patients ≤ 70 years of age, without a history of breast cancer (total n = 944, Supplement enrolled between October 2013 and September 2017 and Table 1) provided by PROFILES [20]. Data of the reference who responded to at least two questionnaires assessing population were matched on age for the three different age PROs including baseline questionnaire (i.e., responders, strata (< 55 years n = 494, 55–65 years n = 247, > 65 years response rates: 70–92%). Patients were excluded when n = 203). 1 3 Breast Cancer Research and Treatment (2019) 175:429–441 431 and non-chemotherapy patients correcting for age, endocrine Statistical analysis treatment, educational level, anxiety and depression, and tumor stage, handling the correlation structures of repeated Patients were categorized according to chemotherapy sta- tus (i.e., non-chemotherapy and chemotherapy group) to measures nested within patients. In sensitivity analysis, we evaluated the effect of hormonal changes on subjective self- observe overall differences in subjective cognitive function. For this purpose, completion dates for questionnaires were reported cognitive functioning. We included pre/perimeno- pausal and postmenopausal status at baseline instead of age compared with the dates of the start of (neo)adjuvant chemo- therapy. Patients were categorized into the chemotherapy to evaluate if the effect of chemotherapy would change, as age is related to subjective cognitive impairment. A sec- group when the start of chemotherapy preceded the date of completing the questionnaire. Therefore, patients were ond sensitivity analysis was performed excluding patients treated with neoadjuvant chemotherapy (i.e., only including only categorized as chemotherapy patient, when chemo- therapy had been started. For example, patients receiving adjuvant chemotherapy patients). We included only patients treated with adjuvant chemotherapy to exclude the effect adjuvant chemotherapy 4 months after baseline were coded as ‘non-chemotherapy group’ at baseline and 3 months, but of more advanced tumor stage and more intense treatment regime, which are most likely patients receiving neoadjuvant as ‘chemotherapy group’ at 6, 12, 18, and 24 months. After 6 months, there were no more shifts from the non-chemo- chemotherapy, on cognitive function. As a measure of clinically meaningful difference, the therapy to the chemotherapy group. All patients that had started chemotherapy, remained in the chemotherapy group. standardized effect size (ES) was calculated (MD divided by the pooled standard deviation of the differences in scores) Next, patients were stratified by age in three groups, patients below 55 years, between 55 and 65 years, and above 65 years and classified as “no effect” (ES 0.2), “small effect” (ES, 0.2–0.4), “medium effect” (ES, 0.5–0.7), and “large effect” of age. Frequencies, proportions, and means with standard deviations for normally distributed variables (and medians (ES ≥ 0.8), according to Cohen [23]. A decline of 10 points on the EORTC QLQ-C30 cognitive function scale was con- with interquartile ranges (IQR) otherwise), were used to describe patient, tumor, and treatment characteristics. sidered clinically meaningful [24, 25]. The level of statistical significance was P < 0.05 and cor- The EORTC QLQ-C30 questionnaire was processed according to the EORTC scoring manual [21]. Cognitive rections for multiple testing were made. Statistical analyses were performed using SPSS Statistics for Windows, version function scores of the EORTC QLQ-C30 range from 0 to 100 with higher scores representing better cognitive func- 23 (IBM Corp, Armonk, NY). tioning [21]. Changes in cognitive function within the non- chemotherapy and chemotherapy within different age strata Results were analyzed with linear mixed-effects models to account for the correlation within subjects between the repeated In total, 1441 patients were enrolled in the cohort between measurements [22]. Linear mixed-effects models tested the presence of an October 2013 and September 2017 (Fig. 1). Of all partici- pants, 979 patients met the inclusion criteria and 715 (73%) age group-by-time-by-treatment (chemotherapy treatment versus no chemotherapy treatment) interactions for subjec- completed 2 or more quality-of-life (QOL) questionnaires. In total, 361 (50%) of the selected patients were treated with tive self-reported cognitive function. The model included a participant-specific random effect. Age, tumor stage, edu- either neoadjuvant (n = 131, 35%) or adjuvant (n = 230, 65%) chemotherapy. cational level, endocrine treatment, anxiety (HADS Anxi- ety score < 8 vs. ≥ 8), and depression (HADS Depression Patients in the chemotherapy group had a median age of 51 (IQR 13) and patients in the non-chemotherapy group had score < 8 vs. ≥ 8) were included as fixed effects to adjust for potential confounding effects. An autoregressive covariance a median age of 60 (IQR 14, Table 1). Patients treated with chemotherapy more often presented with a higher patho- structure of the first order (AR1) was used to define the cor - relations among observations, assuming correlations would logical tumor stage (T2 or T3) and clinical positive lymph nodes and had undergone more extensive treatment (e.g., be greater between measurements that were closer together in time compared with those further apart (i.e., exponential mastectomy, axillary lymph node dissection, and endocrine therapy). decline) [2]. Changes in subjective self-reported cognitive function are presented as the mean differences (MDs) with In total, half of the patients (n = 355) were aged 55 years or younger, 240 patients (33%) between 55 and 65 years and confidence intervals (95% CI), reflecting the difference between patients with chemotherapy and without chemo- 120 patients (17%) were older than 65 years. The majority of patients (n = 227, 64%) in the youngest age group received therapy treatment in the same age category group. The use of advanced modeling approaches allowed us to examine main chemotherapy, whereas 45% (n = 107) of the patients aged between 55 and 65 years and 23% (n = 27) of the > 65 years effects and interactions between age, time, chemotherapy, 1 3 432 Breast Cancer Research and Treatment (2019) 175:429–441 Fig. 1 Flowchart of patient inclusion within the UMBRELLA breast cancer cohort and response rates. Response rates were calcu- lated based on the opportunity patients had to return the questionnaire according to their inclusion date of age, were treated with chemotherapy (Table 2). Educa- reference population (n = 494), cognitive function of both tional level was highest in patients aged < 55 years of age. the chemotherapy and non-chemotherapy patients was lower at all time points. Comparison by treatment group In patients aged between 55 and 65 years, adjusted scores of subjective self-reported cognitive function were Subjective self-reported cognitive functioning, adjusted for significantly lower in those treated with chemotherapy at educational level, age, tumor stage, endocrine therapy, anxi- 3, 12, 18, and 24 months. Mean scores for subjective self- ety, and depression (HADS) of patients treated with chemo- reported cognitive function were significantly worse at all therapy, was significantly worse 3, 6, and 12 months after time intervals except for baseline scores and 6 months (MD the start of radiotherapy compared to non-chemotherapy 6.8, 95% CI 2.2–11.2, MD 8.1, 95% CI 3.1–13.0, MD 7.0, patients. In both groups, subjective self-reported cognitive 95% CI 1.6–12.3, MD 8.0, 95% CI 2.0–13.9, at 3, 12, 18, function deteriorated at 3 months when compared to base- and 24 months, resp., Table 3). The subjective self-reported line scores and improved thereafter with scores comparable cognitive score of the age-matched Dutch reference popula- to baseline scores at 24 months (Fig. 2). tion was higher. Among patients above 65 years of age, the difference in Comparison by age and treatment group cognitive function scores between chemotherapy groups was less pronounced (Fig. 3). At 3 months, patients treated Younger patients (< 55 years) treated with chemotherapy with chemotherapy had lower score compared to non- reported significantly worse subjective self-reported cogni- chemotherapy patients (MD 7.7, 95% CI − 4.3 to 19.7, ES; tive functioning than patients from the same age who did 0.3, Table  3). At 6, 12, 18, and 24 months, self-reported not receive chemotherapy treatment (Fig. 3), after adjusting subjective cognitive function scores were lower, but non- for age, tumor stage, educational level, endocrine therapy, significant, in patients not treated with chemotherapy. anxiety, and depression (HADS) in mixed model analysis The age-matched (> 65 years) Dutch reference population (Table 3). Mean differences (MD) in subjective self-reported showed comparable scores in chemotherapy patients at 18 cognitive functioning scores in patients < 55 years of age, and 24 months. treated with or without chemotherapy were largest at 6, 12, Sensitivity analysis of chemotherapy on subjective self- and 18 months (MD 12.3, 95% CI 8.2–16.3, MD 10.7, 95% reported cognitive function in pre/peri versus postmeno- CI 6.5–14.9, and MD 10.7, 95% CI 5.9–15.4, resp.). Effect pausal patients showed a similar effect of chemotherapy as sizes varied from 0.2 to 0.6, indicating a small/medium in the effect of chemotherapy in patients < 55 years of age effect (Table  2). Compared to the age-matched Dutch and patients between 55 and 65 years of age. The effect of 1 3 Breast Cancer Research and Treatment (2019) 175:429–441 433 Table 1 Demographics and Chemotherapy No chemotherapy disease characteristics of No. of patients (%) No. of patients (%) patients treated with and without chemotherapy in the Total no. of patients 361 (50) 354 (50) first 24 months after enrollment Age in years at inclusion, median (IQR) 51 (13) 60 (14) in the UMBRELLA breast cancer cohort Age category  < 55 years 227 (63) 128 (36)  55–65 years 107 (30) 133 (38)  > 65 years 27 (8) 93 (26) Menopausal status at inclusion  Premenopausal 158 (44) 58 (16)  Peri- or postmenopausal 183 (51) 269 (76)  Unknown 20 (6) 27 (8) Pathological tumor stage  T0 47 (13) 0 (0)  T1 181 (50) 316 (89)  T2 102 (28) 34 (10)  T3 20 (6) 0 (0)  Unknown 11 (3) 4 (1) Course of chemotherapy treatment  Neoadjuvant chemotherapy 135 (37) NA  Adjuvant chemotherapy 226 (63) NA Surgical treatment  Breast-conserving surgery 278 (77) 344 (97)  Mastectomy 83 (23) 10 (3) Most invasive axillary procedure  Sentinel lymph node biopsy 284 (79) 343 (97)  Axillary lymph node dissection 60 (17) 5 (1)  Unknown 17 (5) 6 (2) Estrogen receptor status  Negative 98 (27) 12 (3)  Positive 263 (73) 342 (97)  Unknown 0 (0) 2 (1) HER2 receptor status  Negative 270 (75) 338 (96)  Positive 91 (25) 6 (2)  Unknown 0 (0) 10 (3) Type of chemotherapy  Anthracycline based 328 (91) NA  Non-anthracycline based 34 (9) NA Endocrine treatment  No 109 (30) 218 (62)  Yes 252 (70) 135 (38) Type of radiotherapy  Local radiotherapy 224 (62) 324 (92)  Locoregional radiotherapy 137 (38) 29 (8)  Partial breast 0 (0) 1 (0) Highest educational le vel  Secondary school ± elementary school 79 (22) 96 (27)  Lower vocational education 82 (23) 68 (19)  Community college 111 (31) 84 (24)  University 61 (17) 50 (14)  Unknown 28 (8) 56 (16) 1 3 434 Breast Cancer Research and Treatment (2019) 175:429–441 Table 1 (continued) Chemotherapy No chemotherapy No. of patients (%) No. of patients (%) Working tasks  Physical work 35 (10) 34 (10)  Mental work 145 (40) 105 (30)  Combination of physical and mental work 97 (27) 73 (20)  Unknown 119 (33) 142 (40) HADS score on the Anxiety subscale, mean (SD) 6.2 (3.7) 5.5 (3.4) HADS score on the Depression subscale, mean (SD) 3.7 (3.5) 3.1 (3.3) All patients who received chemotherapy treatment in the first 24 months after enrollment in the UMBRELLA breast cancer cohort were considered chemotherapy patients and contributed to the chemo- therapy group Categories may not sum to total N or 100% because of missing values or rounding IQR interquartile range, NA not applicable, HER2 human epidermal growth factor receptor 2, HADS Hospi- tal Anxiety and Depression scale Neoadjuvant chemotherapy was given in combination with immunotherapy if patients were HER2 recep- tor positive Estrogen receptor positive > 10% Radiotherapy on the breast or chest wall with or without boost on the tumor bed Includes radiotherapy on axillary and/or periclavicular lymph nodes and/or internal mammary nodes Self-repor ted chemotherapy on subjective cognitive function was larg- to be more similar between patients who were exposed to est (ES: 0.4–0.6) in pre/perimenopausal patients (Supple- chemotherapy, not exposed to chemotherapy, and healthy ment Table 3). Sensitivity analysis on the effect of cognitive controls. functioning only including patients who received adjuvant Many cross-sectional and longitudinal studies have found chemotherapy patients shows bigger difference in cognitive evidence for chemotherapy-induced self-reported subjective function in the youngest patient group at baseline (MD 12.9, cognitive changes in patients with breast cancer; however, 95% CI; 3.7–22.0 Supplement Table 4). studies on the effect in different age categories are limited [11, 26–28]. In several studies on risk factors for cognitive decline, age has been found to be a well-established factor Discussion for subjective and objective cognitive deterioration. Even though most studies report that older patients may be more This prospective cohort study showed that self-reported sub- vulnerable to toxic effects of treatment, we did not see an jective cognitive function within the first 2 years after breast (strong) effect of chemotherapy on self-reported subjective cancer was lower in early-stage breast cancer patients treated cognitive function in patients aged over 65 years; however, with chemotherapy compared to patients treated without the number of older patients treated with chemotherapy chemotherapy with significantly lower scores at 3, 6, and was small in this study [2, 29–33]. Longitudinal study in 12 months. The short- and long-term ee ff ct of chemotherapy 112 breast cancer patients from Ahles et al. [32] found that on self-reported subjective cognitive function was most pro- older patients who were exposed to chemotherapy per- nounced in patients younger than 55 years of age. Although formed worse on neuropsychological measures of processing self-reported subjective cognitive functioning of younger speed compared to non-chemotherapy patients and healthy patients slightly improved 3  months after treatment, the controls, i.e., older patients had lower objective cognitive chemotherapy group consistently reported statistically sig- functioning. Lower pretreatment cognitive reserve may nificant worse scores. Up to 2 years after diagnosis, younger make older patients more prone to the negative influence of patients (≤ 65 years) experienced worse self-reported subjec- chemotherapy on brain function, resulting in small cogni- tive cognitive functioning compared to the Dutch reference tive changes in cognition being detectable more easily with population, matched on age reflecting the same age of the neuropsychological tests [34]. Although Ahles et al. [32] two youngest age strata (respectively, < 55 and 55–65 years). corrected for cognitive reserve at baseline, they did not find This was in contrast with older patients (> 65  years), in a significant relation with older age and self-reported subjec- whom self-reported subjective cognitive function was found tive cognitive function, solely chemotherapy treatment was 1 3 Breast Cancer Research and Treatment (2019) 175:429–441 435 Table 2 Demographics and disease characteristics of patients below 55 years of age, between 55 and 65 years of age, and above the age of 65 years, participating in the UMBRELLA breast cancer cohort < 55 years 55–65 years > 65 years No. of patients (%) No. of patients (%) No. of patients (%) Total no. of patients 355 (50) 240 (34) 120 (17) Total no. of patients treated with chemotherapy 227 (64) 107 (45) 27 (23) Age at inclusion, median (IQR) 49 (6) 60 (4) 68 (2)  Menopausal status at inclusion  Premenopausal 213 (60) 3 (1) 0 (0)  Peri- or postmenopausal 99 (28) 233 (97) 120 (100)  Unknown 43 (12) 4 (2) 0 (0) Pathological tumor stage  T0 32 (9) 12 (5) 3 (3)  T1 226 (64) 176 (73) 95 (79)  T2 71 (20) 45 (19) 20 (17)  T3 17 (5) 2 (1) 1 (1)  Unknown 9 (3) 5 (2) 1 (1) Course of chemotherapy treatment  Neoadjuvant chemotherapy 102 (29) 25 (10) 7 (6)  Adjuvant chemotherapy 125 (35) 82 (34) 20 (17) Surgical treatment  Breast-conserving surgery 291 (82) 220 (92) 111 (93)  Mastectomy 64 (18) 20 (8) 9 (8) Most invasive axillary procedure  Sentinel lymph node biopsy 302 (85) 216 (90) 109 (91)  Axillary lymph node dissection 41 (12) 17 (7) 7 (6)  Unknown 12 (3) 7 (3) 4 (3) Estrogen receptor status  Negative 61 (17) 32 (13) 17 (14)  Positive 294 (83) 207 (86) 102 (85)  Unknown 0 (0) 1 (0) 1 (1) HER2 receptor status  Negative 286 (81) 213 (89) 109 (91)  Positive 65 (18) 23 (10) 9 (8)  Unknown 4 (1) 2 (1) 1 (1) Adjuvant chemotherapy treatment  No 230 (65) 158 (66) 100 (83)  Yes 125 (35) 82 (34) 20 (17) Type of chemotherapy  Anthracycline based 207 (58) 99 (41) 23 (19)  Non-anthracycline based 21 (6) 8 (3) 5 (4) Endocrine treatment  No 144 (41) 115 (48) 69 (58)  Yes 211 (59) 125 (52) 51 (43) Type of radiotherapy  Local radiotherapy 254 (72) 194 (81) 100 (83)  Locoregional radiotherapy 101 (28) 45 (19) 20 (17)  Partial breast 0 (0) 1 (0) 0 (0) Highest educational level  Secondary school ± elementary school 66 (19) 77 (32) 32 (27)  Lower vocational education 93 (26) 48 (20) 9 (8)  Community college 109 (31) 73 (30) 13 (11) 1 3 436 Breast Cancer Research and Treatment (2019) 175:429–441 Table 2 (continued) < 55 years 55–65 years > 65 years No. of patients (%) No. of patients (%) No. of patients (%)  University 82 (23) 23 (10) 6 (5)  Unknown 5 (1) 19 (8) 60 (50) HADS score on the Anxiety subscale, mean (SD) 6.2 (3.7) 5.7 (3.7) 5.2 (3.4) HADS score on the Depression subscale, mean (SD) 3.6 (3.5) 3.3 (3.4) 3.1 (3.3) Categories may not sum to total N because of missing values NA not applicable, HER2 human epidermal growth factor receptor 2, HADS Hospital Anxiety and Depression scale Neoadjuvant chemotherapy was given in combination with immunotherapy if patients were HER2 receptor positive Estrogen receptor positive > 10% Radiotherapy on the breast or chest wall with or without boost on the tumor bed Includes radiotherapy on axillary and/or periclavicular lymph nodes and/or internal mammary nodes Self-repor ted Fig. 2 Cognitive functioning. Higher score indicates better cognitive functioning found to be significantly related to self-reported subjective Results showed that self-reported cognitive impairment is cognitive function. Mandelblatt et al. [29] recently showed a substantial and pervasive problem for patients during and that subjective self-reported cognition for survivors (mean after chemotherapy treatment [28]. Younger age and black age 68) exposed to chemotherapy decreases over time, race were associated with problems with perceived cognitive whereas cognition of non-exposed patients did not change; abilities [28]. As cancer treatment, like chemotherapy, may however, the decline among the chemo exposed was non- accelerate the aging process via processes like DNA dam- significant. They found that older patients with genotypes age, it is plausible that the greatest effect of chemotherapy associated with neurodegenerative disease, such as apoli- on cognitive function is detected in younger patients [36]. poprotein E, exposed to chemotherapy showed a clinically Furthermore, it might be that younger patients may detect meaningful decrease in self-reported cognitive function at impairment more readily as they are less likely to attribute 24 months [29]. These results support the idea that not only problems to age-related changes. Also, such subjective cog- aging but also aging phenotypes are associated with lower nitive problems are less likely to be obscured by preexist- self-reported cognitive function. It has been suggested that ing age-related deficits [31]. Younger women may perceive chemotherapy can lead to cancer-related cognitive declines subjective cognitive impairment due to treatment more often through acceleration of aging processes [35]. than older patients because they are more often employed Our study indicates that chemotherapy predominantly has and working, which requires more from their cognition and a substantial detrimental effect on subjective self-reported intellect. cognition in younger women (particularly < 55  years). Besides chemotherapy, other factors may also affect sub- Janelsins et al. compared subjective self-reported cognitive jective cognitive functioning in patients. Also, in women function of patients exposed to chemotherapy (n = 581), ≤ 65 years treated without chemotherapy, cognitive function- mean age of 53 years, to age-matched non-cancer controls. ing was lower compared to healthy age-matched controls. 1 3 Breast Cancer Research and Treatment (2019) 175:429–441 437 Fig. 3 Cognitive function in early-stage breast cancer patients according to chemo- therapy treatment stratified by age categories and compared to the age-matched Dutch reference population. Cogni- tive function was assessed with the cancer questionnaire of the European Organization for Research and Treatment of Cancer (EORTC QLQ- C30). Results are presented as adjusted mean scores accom- panied with 95% confidence intevals. Higher score indicates better cognitive function The impact of a cancer diagnosis itself, surgery, and tox- et al., reported that psychological consequences of a cancer icity due to other treatments may also have effect on the diagnosis and treatment contribute more to cognitive dys- reporting of cognitive symptoms [37]. Furthermore, sub- function than the effects of medication [37]. The feeling of jective cognitive changes are strongly associated with other life disruption and sick leave might be more pronounced in patient-reported outcomes like fatigue, anxiety, and depres- younger patients contributing to greater cognitive impair- sion making it challenging to determine whether cognitive ment compared to older patients [37]. functioning is solely related to underlying brain dysfunc- Sensitivity analysis was performed stratifying by meno- tion [11]. Longitudinal study in 187 breast cancer patients pausal status instead of age. Reason behind this is that evi- with and without chemotherapy exposure from Hermelink dence has emerged suggesting that hormonal changes, alone 1 3 438 Breast Cancer Research and Treatment (2019) 175:429–441 Table 3 Cognitive function in younger (< 55 years), middle-aged (55–65 years), and older (> 65 years) patients with breast cancer assessed with the EORTC QLQ-C30 at baseline, 3, 6, 12, 18, and 24 months after first consultation with the radiation oncologist Age < 55 years Age 55–65 years Age > 65 years Overall P value No chemotherapy n = 128 n = 133 n = 93 (reference group) Chemotherapy n = 277 n = 107 n = 27 a b c a b c a b c MD 95% CI P value ES MD 95% CI P value ES MD 95% CI P value ES Baseline 5.6 1.9–9.3 0.003 0.2 2.9 − 4.1 to 9.9 0.422 0.1 1.7 − 16.6 to 20.1 0.853 0.0 0.017 3 Months 4.4 0.5–8.2 0.024 0.2 6.8 2.2–11.2 0.003 0.3 7.7 − 4.3 to 19.7 0.208 0.3 6 Months 12.3 8.2–16.3 0.000 0.6 4.4 − 0.3 to 9.1 0.072 0.2 − 4.7 − 16.8 to 7.5 0.454 0.2 12 Months 10.7 6.5–14.9 0.000 0.5 8.1 3.1–13.0 0.001 0.4 − 1.8 − 14.2 to 10.5 0.772 0.1 18 Months 10.7 5.9–15.4 0.000 0.5 7.0 1.6–12.3 0.010 0.3 − 9.9 − 23.3 to 3.6 0.151 0.4 24 Months 7.4 1.8–13.2 0.009 0.4 8.0 2.0–13.9 0.009 0.4 − 9.5 − 25.1 to 6.1 0.235 0.3 The differences in mean score (MD) between in the younger, middle-aged, and older patient group are adjusted for endocrine treatment and show the difference in cognitive function between chemotherapy and non-chemotherapy treatment Patient-reported outcomes on cognitive functioning according to EORTC QLQ-C30 Between-group effects were assessed using mixed models including the measurements obtained at baseline and at 3, 6, 12, 18, and 24 months, adjusted for multiple testing, age, tumor stage, endocrine treatment, educational level, anxiety (HADS), and depression (HADS). Patients not exposed to chemotherapy serve as a reference category to calculate mean differences N number of patients, MD mean difference, 95% CI confidence interval, ES effect size Difference in mean score with reference group The P value shown for the difference between no chemotherapy and chemotherapy group Standardized effect size calculated (mean difference divided by the pooled standard deviation) as a measure for minimal clinically important change. Small effect if ES 0.2–0.4, medium 0.5–0.7, large effect > 0.8 The P value shown for the interaction Age-by-time by treatment (no chemotherapy versus chemotherapy) or in combination with chemotherapy and endocrine therapy, genomic testing to guide decisions on withholding chemo- may cause cognitive impairment [29, 38–40]. An observed therapy in selected patients. Furthermore, results of this effect could thus be an age-related effect rather than an inde- study, can help patients and doctors in shared decision mak- pendent effect of systemic treatment per se. Results showed ing about chemotherapy treatment. As many young patients that the ee ff ct of chemotherapy was largest in pre/perimeno - have to return to work, they should be adequately informed pausal women treated with chemotherapy. by physicians about the risks of cognitive decline during and This study adds to the growing literature suggesting after treatment with chemotherapy. that patients experience subjective self-reported cognitive A limitation of our study is the fact that the problems after cancer and/or cancer-related treatments like UMBRELLA cohort was not specifically designed to eval- chemotherapy. The effect of chemotherapy on subjective uate the effect of chemotherapy, and patients were enrolled cognitive function, especially in young patients at 6, 12, in our study at their first consultation with the radiation and 18 months after baseline, emphasizes the importance oncologist instead. In the majority of the patients, this was of adequate selection of early-stage breast cancer patients after breast surgery before the start of radiotherapy treat- whom benefit most from chemotherapy treatment and to ment. However, not all patients were enrolled and com- avoid overtreatment with chemotherapy and its attendant pleted the baseline PRO questionnaires, at the same time toxic effects. In recent years, research has been devoted to in their treatment trajectory (i.e., after surgery, before the the development and validation of genomic tests that can start of radiotherapy). This especially applies for women provide not only prognostic information but perhaps more who had received neoadjuvant chemotherapy treatment importantly can predict response to therapy [41]. Oncotype and were enrolled before starting radiotherapy treatment DX and MammaPrint are well validated and the most widely which were mostly younger patients. Although we tried to used multigene signatures for predicting outcomes in breast correct for these differences in treatment trajectories, we cancer [42, 43]. The use of genomic testing to guide chemo- might have missed early psychological complaints due to therapy treatment has been shown to lead to a reduction in cancer diagnosis and neoadjuvant chemotherapy in some the use of adjuvant chemotherapy in patients with early- patients. Furthermore, the lack of pretreatment subjective stage breast cancer [42]. This highlights the importance of cognitive function, which prevented us from adjusting 1 3 Breast Cancer Research and Treatment (2019) 175:429–441 439 Informed consent Informed consent was obtained from all individual for pretreatment cognitive function, may determine the participants included in the study. amount of cognitive decline. The use of the EORTC Cognitive Functioning alone, to assess subjective self- Open Access This article is distributed under the terms of the Crea- reported cognitive function, may result in an underesti- tive Commons Attribution 4.0 International License (http://creat iveco mation of the extent of an individual’s cognitive symptoms mmons.or g/licenses/b y/4.0/), which permits unrestricted use, distribu- as this EORTC domain encompasses and only captures tion, and reproduction in any medium, provided you give appropriate two aspects of cognition (concentration and memory) credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. [11]. Also, self-assessment of cognitive functioning may be prone to nocebo effects. In other words, information about the association between cognitive difficulties and chemotherapy or endocrine therapy given to chemotherapy References patients and patients who received endocrine therapy (± chemotherapy) might increase the reporting of cognitive 1. Holzel D, Eckel R, Bauerfeind I, Baier B, Beck T, Braun M problems [44]. No formal neuropsychological tests were et al (2017) Improved systemic treatment for early breast can- cer improves cure rates, modifies metastatic pattern and shortens performed to test objective cognitive functioning. Objec- post-metastatic survival: 35-year results from the Munich Cancer tive cognitive functioning might have provided us with Registry. J Cancer Res Clin Oncol 143(9):1701–1712 a different association between chemotherapy and the 2. Ganz PA, Kwan L, Castellon SA, Oppenheim A, Bower JE, Sil- effect on cognition since self-reported cognitive symptoms verman DH et al (2013) Cognitive complaints after breast cancer treatments: examining the relationship with neuropsychological are known to be strongly associated with other patient- test performance. J Natl Cancer Inst 105(11):791–801 reported outcomes like fatigue, anxiety, and depression 3. Boykoff N, Moieni M, Subramanian SK (2009) Confronting [11]. Although we corrected for the use of endocrine ther- chemobrain: an in-depth look at survivors’ reports of impact on apy, we do not have information on compliance to adjuvant work, social networks, and health care response. J Cancer Surviv 3(4):223–232 endocrine treatment. A substantial proportion of patients 4. Winocur G, Johnston I, Castel H (2018) Chemotherapy and cogni- may have not been compliant to endocrine treatment due tion: International Cognition and Cancer Task Force recommen- to side effects. dations for harmonising preclinical research. Cancer Treat Rev In conclusion, patients treated with chemotherapy 69:72–83 5. Early Breast Cancer Trialists’ Collaborative Group (EBCTCG), report more impaired subjective cognitive functioning Peto R, Davies C, Godwin J, Gray R, Pan HC et  al (2012) as compared to those treated without chemotherapy. This Comparisons between different polychemotherapy regimens effect starts shortly after treatment and is seen up until 24 for early breast cancer: meta-analyses of long-term outcome months of follow-up. The association between chemother- among 100,000 women in 123 randomised trials. Lancet 379(9814):432–444 apy and impaired subjective cognitive functioning is most 6. Wefel JS, Vardy J, Ahles T, Schagen SB (2011) International pronounced in younger women, while in older patients dif- Cognition and Cancer Task Force recommendations to harmonise ferences in subjective cognitive function between exposed studies of cognitive function in patients with cancer. Lancet Oncol and non-exposed to chemotherapy were less. These results 12(7):703–708 7. Wefel JS, Saleeba AK, Buzdar AU, Meyers CA (2010) Acute and highlight the careful consideration that is needed in the late onset cognitive dysfunction associated with chemotherapy in current clinical decision making whether chemotherapy women with breast cancer. Cancer 116(14):3348–3356 should be administered to patients with early-stage breast 8. Pullens MJ, De Vries J, Van Warmerdam LJ, Van De Wal MA, cancer. Furthermore, with the emergence of increasing Roukema JA (2013) Chemotherapy and cognitive complaints in women with breast cancer. Psychooncology 22(8):1783–1789 evidence for the efficacy of cognitive training strategies, 9. Schagen SB, Muller MJ, Boogerd W, Rosenbrand RM, van Rhijn there should be more management options available to D, Rodenhuis S et al (2002) Late effects of adjuvant chemotherapy cancer patients to address this important issue. on cognitive function: a follow-up study in breast cancer patients. Ann Oncol 13(9):1387–1397 Acknowledgements We would like to thank Rebecca Stellato for her 10. Koppelmans V, Breteler MM, Boogerd W, Seynaeve C, Gundy C, statistical advice on mixed model analysis. Schagen SB (2012) Neuropsychological performance in survivors of breast cancer more than 20 years after adjuvant chemotherapy. J Clin Oncol 30(10):1080–1086 Compliance with ethical standards 11. Bray VJ, Dhillon HM, Vardy JL (2018) Systematic review of self- reported cognitive function in cancer patients following chemo- Conflict of interest The authors declare that they have no conflict of therapy treatment. J Cancer Surviv 12(4):537–559 interest. 12. Young-Afat DA, van Gils CH, van den Bongard HJGD, Verkooi- jen HM, UMBRELLA Study Group (2017) The Utrecht cohort Ethical approval All procedures performed in studies involving human for Multiple BREast cancer intervention studies and Long-term participants were in accordance with the ethical standards of the insti- evaLuAtion (UMBRELLA): objectives, design, and baseline tutional and/or national research committee and with the 1964 Helsinki results. Breast Cancer Res Treat 164(2):445–450 declaration and its later amendments or comparable ethical standards. 1 3 440 Breast Cancer Research and Treatment (2019) 175:429–441 13. Risicoprofilering Richtlijn Mammacarcinoom (2012) Kwaliteit- of breast cancer: CALGB 369901 (Alliance). Cancer 2016 sinstituut voor de gezondheidszorg CBO VvIK, pp 81–83 122(22):3555–3563 14. Netherlands Cancer Registry (2018) http://www.cijfe rsove rkank 31. Debess J, Riis JO, Engebjerg MC, Ewertz M (2010) Cogni- er.nl. Accessed May 2018 tive function after adjuvant treatment for early breast cancer: a 15. Aaronson NK, Ahmedzai S, Bergman B, Bullinger M, Cull A, population-based longitudinal study. Breast Cancer Res Treat Duez NJ et al (1993) The European Organization for Research 121(1):91–100 and Treatment of Cancer QLQ-C30: a quality-of-life instrument 32. Ahles TA, Saykin AJ, McDonald BC, Li Y, Furstenberg CT, for use in international clinical trials in oncology. J Natl Cancer Hanscom BS et  al (2010) Longitudinal assessment of cogni- Inst 85(5):365–376 tive changes associated with adjuvant treatment for breast 16. Spinhoven P, Ormel J, Sloekers PP, Kempen GI, Speckens AE, cancer: impact of age and cognitive reserve. J Clin Oncol Van Hemert AM (1997) A validation study of the Hospital Anxi- 28(29):4434–4440 ety and Depression Scale (HADS) in different groups of Dutch 33. Hurria A, Hurria A, Zuckerman E, Panageas KS, Fornier M, subjects. Psychol Med;27(2):363–370 D’Andrea G et al (2006) A prospective, longitudinal study of the 17. Zigmond AS, Snaith RP (1983) The hospital anxiety and depres- functional status and quality of life of older patients with breast sion scale. Acta Psychiatr Scand 67(6):361–370 cancer receiving adjuvant chemotherapy. J Am Geriatr Soc 18. Smith AB, Selby PJ, Velikova G, Stark D, Wright EP, Gould A 54(7):1119–1124 et al (2002) Factor analysis of the Hospital Anxiety and Depres- 34. Healey MK, Campbell KL, Hasher L (2008) Cognitive aging and sion Scale from a large cancer population. Psychol Psychother increased distractibility: costs and potential benefits. Prog Brain 75(Pt 2):165–176 Res 169:353–363 19. Snaith RP (2003) The Hospital Anxiety And Depression Scale. 35. Vega JN, Dumas J, Newhouse PA (2018) Self-reported chemother- Health Qual Life Outcomes 1:29–7525 apy-related cognitive impairment compared with cognitive com- 20. van de Poll-Franse LV, Horevoorts N, van Eenbergen M, Denollet plaints following menopause. Psychooncology 27(9):2198–2205 J, Roukema JA, Aaronson NK et al (2011) The Patient Reported 36. Ahles TA, Root JC, Ryan EL (2012) Cancer- and cancer treat- Outcomes Following Initial treatment and Long term Evaluation ment-associated cognitive change: an update on the state of the of Survivorship registry: scope, rationale and design of an infra- science. J Clin Oncol 30(30):3675–3686 structure for the study of physical and psychosocial outcomes in 37. Hermelink K, Buhner M, Sckopke P, Neufeld F, Kaste J, Voigt V cancer survivorship cohorts. Eur J Cancer 47(14):2188–2194 et al (2017) Chemotherapy and post-traumatic stress in the cau- 21. Fayers PM, Aaronson NK (2001) The EORTC QLQ-C30 scoring sation of cognitive dysfunction in breast cancer patients. J Natl manual, 3rd edn. European Organisation for Research and Treat- Cancer Inst. https ://doi.org/10.1093/jnci/djx05 7 ment of Cancer, Brussels 38. Mandelblatt JS, Jacobsen PB, Ahles T (2014) Cognitive effects of 22. Bonnetain F, Fiteni F, Efficace F, Anota A (2016)Statistical chal- cancer systemic therapy: implications for the care of older patients lenges in the analysis of health-related quality of life in cancer and survivors. J Clin Oncol 32(24):2617–2626 clinical trials. J Clin Oncol 34(16):1953–1956 39. Schilder CM, Eggens PC, Seynaeve C, Linn SC, Boogerd W, 23. ’Cohen J (1988) Statistical power analysis for the behavioral sci- Gundy CM et  al (2009) Neuropsychological functioning in ences, 2nd edn. Lawrence Erlbaum, Hillsdale postmenopausal breast cancer patients treated with tamoxifen 24. Bedard G, Zeng L, Lam H, Cella D, Zhang L, Lauzon N et al or exemestane after AC-chemotherapy: cross-sectional findings (2012) Meaningful change in oncology quality-of-life instru- from the neuropsychological TEAM-side study. Acta Oncol ments: a systematic literature review. Expert Rev Pharmacoecon 48(1):76–85 Outcomes Res 12(4):475–483 40. Mandelblatt JS, Clapp JD, Luta G, Faul LA, Tallarico MD, 25. Raman S, Ding K, Chow E, Meyer RM, van der Linden YM, McClendon TD et  al (2016) Long-term trajectories of self- Roos D et al (2018) Minimal clinically important differences in reported cognitive function in a cohort of older survivors of breast the EORTC QLQ-C30 and brief pain inventory in patients under- cancer: CALGB 369901 (Alliance). Cancer 122(22):3555–3563 going re-irradiation for painful bone metastases. Qual Life Res 22 27(4):1089–1098 41. Nicolini A, Ferrari P, Duffy MJ (2018) Prognostic and predic- 26. Jung MS, Zhang M, Askren MK, Berman MG, Peltier S, Hayes tive biomarkers in breast cancer: past, present and future. Semin DF et al (2017) Cognitive dysfunction and symptom burden in Cancer Biol 52(Pt 1):56–73 women treated for breast cancer: a prospective behavioral and 42. Cardoso F, van’t Veer LJ, Bogaerts J, Slaets L, Viale G, Delaloge fMRI analysis. Brain Imaging Behav 11(1):86–97 S et al (2016) 70-Gene signature as an aid to treatment decisions 27. Quesnel C, Savard J, Ivers H (2009) Cognitive impairments asso- in early-stage breast cancer. N Engl J Med 375(8):717–729 ciated with breast cancer treatments: results from a longitudinal 43. Sparano JA, Gray RJ, Makower DF, Pritchard KI, Albain KS, study. Breast Cancer Res Treat 116(1):113–123 Hayes DF et al (2015) Prospective validation of a 21-gene expres- 28. Janelsins MC, Heckler CE, Peppone LJ, Kamen C, Mustian KM, sion assay in breast cancer. N Engl J Med 373(21):2005–2014 Mohile SG et  al (2017) Cognitive complaints in survivors of 44. Jacobs W, Das E, Schagen SB (2017) Increased cognitive problem breast cancer after chemotherapy compared with age-matched reporting after information about chemotherapy-induced cognitive controls: an analysis from a nationwide, multicenter, prospective decline: The moderating role of stigma consciousness. Psychol longitudinal study. J Clin Oncol 35(5):506–514 Health 32(1):78–93 29. Mandelblatt JS, Small BJ, Luta G, Hurria A, Jim H, McDonald BC et al (2018) Cancer-related cognitive outcomes among older Publisher’s Note Springer Nature remains neutral with regard to breast cancer survivors in the thinking and living with cancer jurisdictional claims in published maps and institutional affiliations. study. J Clin Oncol. https ://doi.org/10.1200/JCO.18.00140 30. Mandelblatt JS, Clapp JD, Luta G, Faul LA, Tallarico MD, McClendon TD et  al (2016) Long-term trajectories of self- reported cognitive function in a cohort of older survivors 1 3 Breast Cancer Research and Treatment (2019) 175:429–441 441 Affiliations 1 2 1,3 4 4 5 M. L. Gregorowitsch  · A. Ghedri  · D. A. Young‑Afat  · R. Bijlsma  · I. O. Baas  · C. van Schaikv ‑ an de Mheen  · 6 7 8 1 2,9 M. J. Agterof  · E. Göker  · D. ten Bokkel Huinink  · H. J. G. D. van den Bongard  · H. M. Verkooijen  · UMBRELLA study group 1 6 Department of Radiation Oncology, University Department of Medical Oncology, St. Antonius, Nieuwegein, Medical Center (UMC) Utrecht, Heidelberglaan 100, The Netherlands 3584 CX Utrecht, The Netherlands Department of Medical Oncology, Alexander Monro Clinics, University of Utrecht, Utrecht, The Netherlands Bilthoven, The Netherlands 3 8 Department of Epidemiology, Julius Center for Health Department of Medical Oncology, Diakonessenhuis, Utrecht, Sciences and Primary Care, University Medical Center, The Netherlands Utrecht, The Netherlands Imaging Division, University Medical Center Utrecht, Department of Medical Oncology, University Medical Center Utrecht, The Netherlands Utrecht, Utrecht, The Netherlands Department of Medical Oncology, Meander Medical Center, Amersfoort, The Netherlands 1 3

Journal

Breast Cancer Research and TreatmentSpringer Journals

Published: Feb 12, 2019

There are no references for this article.