466

‘Chemobrain’ in Breast Carcinoma?

A Prologue

Jeffrey S. Wefel, Ph.D.1 BACKGROUND. Chemotherapy-induced cognitive dysfunction in patients with

Renato Lenzi, M.D.2 breast carcinoma has been described previously. However, those studies only
Richard Theriault, D.O.3 assessed patients’ postchemotherapy cognitive functioning and were not able to
Aman U. Buzdar, M.D.3 determine the relation between cognitive function and other treatments, such as
Scott Cruickshank, M.A.4 surgery and radiotherapy, that often precede systemic chemotherapy.
Christina A. Meyers, Ph.D.1 METHODS. Eighty-four women with breast carcinoma underwent a comprehensive
neuropsychologic evaluation before receiving adjuvant therapy for nonmetastatic
1
Department of Neuro-Oncology, The University of primary breast carcinoma.
Texas M. D. Anderson Cancer Center, Houston, RESULTS. Before the start of systemic therapy, 35% of women in the current cohort
Texas. exhibited cognitive impairment. Verbal learning (18%) and memory function (25%)
2
Department of Gastrointestinal Oncology, The were impaired significantly more frequently relative to normative expectations.
University of Texas M. D. Anderson Cancer Center, Although the impairments were not significant in the women who were examined,
Houston, Texas. nonverbal memory (17%), psychomotor processing speed and attention (13%),
3
Department of Breast Medical Oncology, The confrontational naming (13%), visuoconstruction (13%), and upper-extremity fine
University of Texas M. D. Anderson Cancer Center, motor dexterity (12%) were impaired more frequently than was expected. Affective
Houston, Texas. distress was related significantly to cognitive impairment (Pearson chi-square
4
Scott Cruickshank and Associates, Santa Bar- ⫽ 9.90; P ⫽ 0.002). Given the conservative statistical approach employed, extent of
bara, California. surgery, hormone replacement therapy history, and current menopausal status
failed to achieve statistical significance, but these variables did exhibit provocative
trends with respect to cognitive impairment.
CONCLUSIONS. Cognitive impairment frequently is observed before the adminis-
tration of systemic chemotherapy. Thus, investigations purporting to measure
chemotherapy-induced cognitive dysfunction must employ study designs that
incorporate prechemotherapy baseline assessments to accurately detect changes
in cognitive function that are attributable to chemotherapy. Cancer 2004;101:
Supported in part by a Dissertation Research Grant 466 –75. © 2004 American Cancer Society.
(to J.S.W.) from the Susan G. Komen Foundation.

The authors thank Gabriel N. Hortobagyi, M.D. KEYWORDS: cognition disorders, affect, breast neoplasms, drug therapy.
(Nellie B. Connally Chair in Breast Cancer, Depart-
ment of Breast Medical Oncology, The University of
Texas M. D. Anderson Cancer Center), for the
participation of his department in ongoing collab-
orations; and Lori Smythe (Research Coordinator,
C ancer and the therapeutic modalities used to treat cancer have
been associated with alterations in patients’ cognitive, emotional,
and neurologic functioning.1 It is widely accepted that disease involv-
Department of Neuro-Oncology, The University of ing the central nervous system (CNS) and treatments targeting the
Texas M. D. Anderson Cancer Center) and Carol CNS (e.g., radiotherapy) may have adverse effects on cognitive func-
Ann Long, Ph.D. (Director, Medical Affairs Oncol-
tioning. There is growing awareness that malignant disease outside of
ogy, Amgen, Inc.), for their technical assistance.
the CNS and the treatment of such malignancies with biologic, im-
Address for reprints: Christina A. Meyers, Ph.D., De- munologic, and/or hormonal techniques also may result in alter-
partment of Neuro-Oncology (Box 431), The Univer- ations in patients’ mental status.2
sity of Texas M. D. Anderson Cancer Center, 1515 Breast carcinoma is the second most frequent cause of malignancy-
Holcombe Boulevard, Houston, TX 77030; Fax: (713)
related death among women in the United States. It is estimated that
794-4999; E-mail: [email protected]
nearly one in nine women will develop some form of breast carci-
Received February 23, 2004; revision received noma in their lifetime.3 Despite optimal surgical treatment for breast
April 22, 2004; accepted April 30, 2004. carcinoma, residual tumor cells may remain and either fail to be

© 2004 American Cancer Society

DOI 10.1002/cncr.20393
Published online 21 June 2004 in Wiley InterScience (www.interscience.wiley.com).
 Chemobrain in Breast Carcinoma?/Wefel et al. 467

removed at the surgical site or disseminate to other ing ‘chemobrain’. This finding exemplifies emerging
areas of the body. For this reason, adjuvant therapy public awareness and concern surrounding the issue
(e.g., radiotherapy, chemotherapy, hormonal thera- of chemobrain and underscores the importance of
pies) has become a standard component of care for performing methodologically rigorous studies to help
many women with breast carcinoma. The use of ad- assess the frequency and nature, as well as the poten-
juvant chemotherapy is widely recognized as an im- tial mechanisms, of chemotherapy-induced neurotox-
portant component of multimodal treatment. It was icity.
found that adjuvant chemotherapy increased disease- The potential neurotoxic effects of adjuvant che-
free and overall survival for many patients with breast motherapeutic treatments are a serious concern for
carcinoma.4,5 However, chemotherapy has been asso- many women with breast carcinoma, and well de-
ciated with adverse effects, such as nausea, peripheral signed investigations to help elucidate the potential
neuropathy, and encephalopathy.6,7 The relation be- risks and benefits of such treatments are necessary.
tween chemotherapy and cognitive functioning is an Characterizing the neuropsychologic profiles of
area of active interest in both clinical and research women with breast carcinoma who have not previ-
settings. ously received chemotherapy will enhance our under-
A clear understanding of the relation between standing of the multiple factors that may contribute to
chemotherapeutic treatments and cognitive function- cognitive dysfunction in this population. In addition,
ing is critical, because there is a growing group of these data should reinforce the importance of using
malignancy survivors who often want to return to their prospective trials to assess change in cognitive func-
former occupational, scholastic, or familial activities. tion secondary to chemotherapy, as the use of retro-
In many patients, the integrity of cognitive function- spective analytic techniques is a methodologic limita-
ing is critical to the fulfillment of this goal. Recently, a tion of all published trials to date. To this end, we have
number of studies of women with breast carcinoma examined data that were generated by three separate
have described changes in cognitive functioning that prospective clinical research trials conducted at the
are purportedly associated with chemotherapy. Sev- University of Texas M. D. Anderson Cancer Center
eral retrospective studies have examined cognitive examining the effects of hormonal treatments (2001–
functioning months to years after chemotherapy and 2003) and chemotherapeutic treatments (1992–1995
concluded that chemotherapy in patients with breast and 1994 –998) on cognitive functioning in a popula-
carcinoma is associated with persistent cognitive def- tion of patients with nonmetastatic breast carcinoma.
icits after the completion of treatment.8 –13 The term In each investigation, patients received a comprehen-
chemobrain has been coined in the lay press to de- sive neuropsychologic evaluation before the start of
scribe the cognitive decline associated with chemo- systemic chemotherapy or hormonal therapy.
therapy. However, to examine the cognitive effects of
a therapeutic intervention rigorously, it is of para- MATERIALS AND METHODS
mount importance that investigators obtain a pre- Patients were recruited as part of three institutional
treatment measure of neuropsychologic functioning. review board–approved research protocols that were
Only in this manner can investigators determine initiated collaboratively by the Neuropsychology Sec-
whether the observed performance in subsequent tion and the Department of Breast Medical Oncology
evaluations (i.e., postchemotherapy) represents a at the University of Texas M. D. Anderson Cancer
change in functioning secondary to treatment or Center. However, neuropsychologic outcome was a
whether the observed deficits were present before the nonfunded endpoint in two of those trials. Patient
administration of the therapeutic agent(s). Remark- enrollment and collection of data on this endpoint
ably, this critical assessment time point has been ab- were terminated early due to funding issues; thus, the
sent in all published neuropsychologic studies to date. overall patient pool for the analysis of this endpoint
Each of those studies concluded that there was clini- was smaller than the patient pool that was available in
cally significant cognitive impairment associated with the primary medical trials. All patients who met initial
chemotherapy. However, conclusive support for this screening criteria and consented to participate in the
association requires longitudinal data that include ob- trials were registered consecutively into a protocol for
jective measurement of patients’ prechemotherapy evaluation. All patients had a diagnosis of primary
cognitive functioning. Anecdotally, several patients at breast carcinoma and no evidence of metastatic dis-
our institution (The University of Texas M. D. Ander- ease, were age ⱖ 18 years, had completed ⱖ 8 years of
son Cancer Center, Houston, TX) have expressed res- formal education, and spoke fluent English. No pa-
ervations regarding the receipt of agents with known tient had a history of other primary malignancy or a
beneficial effects secondary to concerns about acquir- previous or current neurologic or psychiatric disorder,
468 CANCER August 1, 2004 / Volume 101 / Number 3

and no patient used substances that were active in the TABLE 1

CNS and were believed to affect cognition (e.g., nar- Neuropsychologic Tests Grouped by Principle Cognitive Domain
cotics, antiemetics, or steroids) ⱕ 1 week before test-
Domain/test namea Test abbreviation
ing. Ninety-seven patients met the initial screening
criteria for participation. Eighty-four patients with a Attention
diagnosis of Stage I breast carcinoma (T1N0M0; n WAIS-R15/WAIS-III16 Digit Span WAIS-R/III DSpan
⫽ 32), Stage II breast carcinoma (T2N0M0; n ⫽ 41), or WAIS-R Digit Symbol WAIS-R DSymbol
WAIS-R Arithmetic WAIS-R Arith
Stage IIIA breast carcinoma (T2N2M0; n ⫽ 11) quali- WAIS-III Letter-Number Sequencing WAIS-III LN
fied for study entry and underwent neuropsychologic WMS-III17 Mental Control WMS-III MC
evaluation before beginning chemotherapy. Thirteen Trail Making Test18 Part A TMTA
patients did not meet the inclusion criteria and were Memory
HVLT19 Trials 1–3 HVLT T1–T3
not part of the subsequent analyses. Of these, seven
HVLT Recognition Discriminability HVLT DISCRIM
patients stated that they did not have time to partici- VSRT20 Long-Term Storage VSRT LTS
pate, two patients had educational limitations that VSRT Delayed Recall VSRT DR
precluded valid assessment, two patients had histories NVSRT21 Long-Term Storage NVSRT LTS
NVSRT Delayed Recall NVSRT DR
of major depression, one patient had a history of sub-
ROCFT22 Delayed Recall ROCFT DR
stance abuse, and one patient had a history of previ- Language
ous breast carcinoma. MAE23 Controlled Oral Word Association COWA
Of the 84 patients who were included in the anal- Boston Naming Test24 BNT
yses, 50% (n ⫽ 42) had undergone surgical resection MAE Sequential Commands MAE SC
Executive
for primary breast carcinoma, and 50% (n ⫽ 42) had
Trail Making Test Part B TMTB
undergone core-needle biopsy before the neuropsy- Category Test25 CT
chologic evaluation in association with their partici- WAIS-III Similarities WAIS-III Sim
pation in a trial that involved neoadjuvant chemother- Visuospatial
WAIS-R Block Design WAIS-R BD
apy after biopsy for pathologic verification.
ROCFT Copy ROCFT Copy
Approximately 12% of all patients (n ⫽ 10) had re- Judgment of Line Orientation26 JLO
ceived adjuvant radiotherapy to the chest wall with or Motor
without the inclusion of regional lymphatic chains Grip Strength (dominant hand)27 GRIP (dominant)
before neuropsychologic evaluation. The mean (⫾ Grooved Pegboard (dominant hand)28 PEG (dominant)
Depression
standard deviation [SD]) patient age was 50.4 years (⫾ Beck Depression Inventory29 BDI
9.1 years). On average, patients had completed 14 Beck Depression Inventory, Second Edition30 BDI-2
years of education (⫾ 2.6). Ethnically, 69 patients MMPI (Depression scale)31 MMPI (Scale 2)
(82%) were Caucasian, 8 (10%) were Hispanic, 6 (7%) Anxiety
State-Trait Anxiety Inventory—State score32 STAIS
were African American, and 1 (1%) was Asian. Data on
Beck Anxiety Inventory33 BAI
menopausal status (based on review of patients’ med- MMPI (Psychasthenia scale) MMPI (Scale 7)
ical records) were available for all patients: approxi-
mately 42% (n ⫽ 35) were naturally postmenopausal, WAIS-R/III: Wechsler Adult Intelligence Scale—Revised or Third Edition; WMS-III: Wechsler Memory
15% (n ⫽ 13) had a history of surgically induced Scale—Third Edition; HVLT: Hopkins Verbal Learning Test; VSRT: Verbal Selective Reminding Test;
NVSRT: Nonverbal Selective Reminding Test; ROCFT: Rey–Osterreith Complex Figure Test; MAE:
menopause, 2% (n ⫽ 2) were perimenopausal, and Multilingual Aphasia Examination; MMPI: Minnesota Multiphasic Personality Inventory.
40% (n ⫽ 34) were premenopausal at the time of a
Superscript numerals refer to the list of references.
neuropsychologic evaluation. Data on prior use of
hormone replacement therapy (HRT) were available
for 67 of 84 patients: Thirty patients (45%) had previ- measures, which required approximately 40 –120 min-
ously used HRT, whereas 37 (55%) had never used utes to be completed.
HRT. The median time between neuropsychologic Patients’ raw cognitive test score performances
evaluation and last use of HRT was 52 days (range, 4 were converted to standardized scores (z scores: mean
days to 23.9 years). ⫾ SD, 0 ⫾ 1) based on published normative data to
Due to the nature of the research studies and facilitate comparisons between measures. An Overall
changes in clinical practice, patients received a variety Cognitive Function Index (OCFI) for each patient’s
of cognitive tests. To facilitate interpretation, these test performance was calculated. Impaired OCFI
tests were grouped into domains based on previous (OCFI-I) was ascertained using a 2-step criterion: 1) if
neuropsychologic research14 (Table 1). All patients un- a patient had multiple test performance z scores
derwent an assessment that included 5–14 different ⱕ ⫺1.5, then she was considered to have impaired
 Chemobrain in Breast Carcinoma?/Wefel et al. 469

functioning (OCFI-I); 2) if a patient had only 1 test that tical tests performed. Cognitive tests that appeared to
met the aforementioned criteria (i.e., z score ⱕ ⫺1.5), be most sensitive were selected as marker variables
then that single z score had to be ⱕ ⫺2.0 for the (i.e., the Trailmaking Test Part A [TMTA]; the Verbal
patient to be classified as having impaired function- Selective Reminding Test, Long-Term Storage [VSRT
ing. This two-step approach was used to minimize the LTS]; the VSRT Delayed Recall [DR]; the Nonverbal
number of false-positive errors due to multiple tests Selective Reminding Test [NVSRT] DR; the Boston
and to determine the frequency of impairment rather Naming Test [BNT]; the Trailmaking Test Part B
than the frequency of low performance levels. In a [TMTB]; the Rey–Osterreith Complex Figure Test
normal, healthy population, approximately 7% and 2% [ROCFT] Copy; and the Grooved Pegboard Test [dom-
of the population would have scores ⬍ ⫺1.5 and inant hand]). Given the large number of statistical
⬍ ⫺2.0, respectively, by chance alone. If a patient did tests performed, ␣, the cutoff value for determining
not exhibit impaired performance, then the OCFI clas- statistical significance, was set at 0.01.
sification was not impaired (OCFI-NI).
For each cognitive measure, a binomial test was RESULTS
conducted to determine whether the frequency of im- Using the classification criteria described above, ap-
pairment observed in the study group differed from proximately 35% (29 of 84 patients; binomial test: P
normative expectations. To correct for multiple com- ⬍ 0.001) received a classification of OCFI-I before the
parisons, ␣, the cutoff value for determining statistical initiation of adjuvant chemotherapy. Of the patients
significance, was set at 0.01. The overall frequency of who had impaired functioning, approximately 31% (9
cognitive impairment based on the OCFI and the fre- of 29) exhibited impairment on 1 test, whereas ap-
quency of impairment within each domain were then proximately 69% (20 of 29) exhibited impairment on 2
calculated. or more tests. The frequency of cognitive impairment
Self-reports of anxious and depressive symptoms within each domain varied as a function of the test
warranted a classification of clinically significant dis- that was used to make the measurement (Table 2).
tress–impaired (CSD-I) if the reports exceeded the lim- Women with breast carcinoma exhibited impaired
its recommended by the measures’ manuals. Specifi- verbal learning (VSRT LTS; Binomial test: P ⬍ 0.01)
cally, patients with raw scores ⱖ 18 on the Beck and verbal memory (VSRT DR; Binomial test: P ⬍ 0.01)
Depression Inventory (BDI), the BDI Second Edition significantly more frequently relative to normative ex-
(BDI-2), and the Beck Anxiety Inventory (BAI) received pectations. In addition, 1 measure of psychomotor
a classification of CSD-I. Similarly, patients who had processing speed and attention approached signifi-
scores ⱖ the 95th percentile on the ‘State’ subscale of cance (TMTA; P ⫽ 0.017). Several other measures of
the State-Trait Anxiety Inventory (STAIS) received a cognitive functioning (NVSRT DR, BNT, ROCFT Copy,
classification of CSD-I. Patients with scores lower than and PEG [dominant hand]) that revealed levels of im-
these specified values received a classification of not pairment similar to those revealed by the TMTA failed
clinically significantly distressed (CSD-NI). All patients’ to approach significance, probably due to small sam-
Minnesota Multiphasic Personality Inventory (MMPI) ple sizes.
profiles yielded adequate validity scale indices to be In the current cohort, 26% of patients (20 of 78)
considered for further clinical scale interpretation.34 reported symptoms of anxiety and/or depression that
Scale 2 of the MMPI measures depression through met the criteria for a classification of CSD-I. Chi-
true/false statements regarding affective symptoms square analysis demonstrated that patients who were
(e.g., poor morale), cognitive symptoms (e.g., hope- classified as having reported significant affective dis-
lessness), and physical symptoms (e.g., sleep distur- tress (i.e., CSD-I) were significantly more likely to be
bances). Scale 7 of the MMPI measures anxiety across cognitively impaired (i.e., OCFI-I; chi-square ⫽ 9.90; P
the domains of neuroticism, anxiety, withdrawal, poor ⱕ 0.005; Fig. 1). Among patients with a classification of
concentration, agitation, psychotic tendencies, and OCFI-I who completed mood assessment (n ⫽ 28),
poor physical health. Patients with T scores ⱖ 70 on ⬍ 50% (13 of 28) received a classification of CSD-I.
Scale 2 or Scale 7 of the MMPI received a classification Scores on the self-report measures of affective distress
of CSD-I; patients with scores below this cutoff re- for patients with a classification of CSD-I were as
ceived a classification of CSD-NI. follows: BDI raw score range, 24 –28; BDI-II raw score,
Pearson correlations between mood measures 20; BAI raw score range, 19 –29; STAIS percentile
(i.e., BDI, BDI-2, and BAI raw scores; STAIS percen- range, ⬍ 1 to 6; MMPI Scale 2 T score range, 71– 80;
tiles; and MMPI Scale 2 and Scale 7 T scores) and and MMPI Scale 7 T score, 74. Overall, these scores are
standard scores for marker variables from each do- consistent with mild-to-moderate ranges of affective
main were analyzed to decrease the number of statis- distress, and no patient received a psychiatric diagno-
470 CANCER August 1, 2004 / Volume 101 / Number 3

TABLE 2 TABLE 3
Mean Scores and Impairment Frequencies Before Chemotherapy on Coefficients for Correlations between Mood Measures
Tests Grouped by Principle Cognitive Domain and Cognitive Marker Variables

No. of Mood measurea

Cognitive domain patients tested Mean (SD) Impaired (%)
MMPI MMPI
Attention Cognitive test BDI BDI-2 BAI STAIS (Scale 2) (Scale 7)
WAIS-R/III DSpana 84 10.0 (3.0) 5
WAIS-R DSymbola 60 11.7 (2.5) 2 Attention
WAIS-R Aritha 18 10.2 (2.6) 6 TMTA 0.20 NA NA ⫺0.30 0.31 ⫺0.11
WAIS-III LNa 24 11.7 (3.3) 4 Learning
WAIS-III MCa 24 11.2 (2.6) 0 VSRT LTS NA NA NA NA ⫺0.07 0.01
TMTAb 84 ⫺0.11 (1.37) 13 Memory
Learning VSRT DR NA NA NA NA ⫺0.16 ⫺0.14
HVLT T1–T3b 42 ⫺0.09 (0.90) 10 NVSRT DR NA NA NA NA ⫺0.38 ⫺0.15
VSRT LTSb 40 ⫺0.27 (1.20) 18c Language
NVSRT LTSb 18 0.11 (0.83) 6 BNT NA ⫺0.44 ⫺0.44 NA NA NA
Memory Executive
HVLT DISCRIMb 42 0.29 (0.71) 2 TMTB 0.04 NA NA ⫺0.42b ⫺0.25 ⫺0.24
VSRT DRb 40 ⫺0.54 (1.80) 25d Visuospatial
NVSRT DRb 18 0.00 (1.04) 17 ROCFT Copy NA ⫺0.32 ⫺0.30 NA NA NA
ROCFT DRb 24 0.13 (0.82) 4 Motor
Language PEG NA NA NA NA ⫺0.20 ⫺0.36
Expressive
COWAb 84 6 BDI: Beck Depression Inventory; BDI-2: Beck Depression Inventory—Second Edition; BAI: Beck Anxiety
BNTb 24 13 Inventory; STAIS: State-Trait Anxiety Inventory—State scale; MMPI: Minnesota Multiphasic Personality
Receptive Inventory; TMTA: Trail Making Test, Part A; VSRT: Verbal Selective Reminding Test; LTS: Long-Term
MAE SCb 24 0 Storage; DR: Delayed Recall; NVSRT: Nonverbal Selective Reminding Test; BNT: Boston Naming Test;
Executive function TMTB: Trail Making Test, Part B; ROCFT: Rey–Osterreith Complex Figure Test; PEG: Grooved Pegboard
TMTBb 84 0.13 (1.90) 6 Test; NA: not administered (due to variation between protocols, neither measure was administered).
CTb 41 0.22 (0.91) 0 a
Metric: raw score for BDI, BDI-2, and BAI; percentile for STAIS; T score for MMPI Scale 2 and Scale
WAIS-III Sima 42 12.1 (3.4) 5 7; and z score for cognitive tests.
Visuospatial b
Pearson correlation: P ⬍ 0.01.
Perceptual
JLOb 24 0.91 (0.38) 0
Perceptual-motor TABLE 4
WAIS-R BDa 18 11.4 (2.9) 0 Frequency of Clinically Significant Before Chemotherapy Distress
ROCFT Copyb 24 0.31 (1.34) 13 on Self-Report Measures by Domain
Motor
GRIP (dominant)b 18 0.11 (0.98) 0 Psychologic domain No. of patients assessed CSD-I (%)
PEG (dominant)b 42 0.02 (1.14) 12
Depression
SD: standard deviation; WAIS-R/III: Wechsler Adult Intelligence Scale—Revised or Third Edition;
MMPI (Scale 2) 17 35
DSpan: Digit Span; DSymbol: Digit Symbol; Arith: Arithmetic; LN: Letter-Number Sequencing; WMS-III
BDI-2 24 8
MC: Wechsler Memory Scale—Third Edition, Mental Control; TMTA: Trail Making Test, Part A; HVLT
BDI 42 4
T1–T3: Hopkins Verbal Learning Test (LVLT) Trials 1–3; VSRT: Verbal Selective Reminding Test; LTS:
Anxiety
Long-Term Storage; NVSRT: Nonverbal Selective Reminding Test; DISCRIM: Recognition Discrim-
MMPI (Scale 7) 17 35
inability; DR: Delayed Recall; ROCFT: Rey–Osterreith Complex Figure Test; COWA: Multilingual Aphasia
STAIS 42 32
Examination (MAE) Controlled Oral Word Association; BNT: Boston Naming Test; SC Sequential
BAI 24 8
Commands; TMTB: Trail Making Test, Part B; CT: Category Test; Sim: Similarities; JLO: Judgment of
Lone Orientation; BD: block design; GRIP (dominant): Grip Strength (dominant hand); PEG (dominant):
CSD-I: clinically significant distress; MMPI: Minnesota Multiphasic Personality Inventory; BDI-2: Beck
Grooved Pegboard (dominant hand).
a
Depression Inventory—Second Edition; BDI: Beck Depression Inventory; STAIS: State-Trait Anxiety
Scaled scores: mean, 10; standard deviation, 3.
b
Inventory—State scale; BAI: Beck Anxiety Inventory.
z scores: mean, 0; standard deviation, 1.
c
Binomial test (P ⬍ 0.01).
d
Binomial test (P ⬍ 0.001).
sociations between mood and cognitive function were
detected. Differences in the frequency of distress
sis of major depression or an anxiety disorder. Analy- within the domains of mood function also varied,
sis of Pearson correlations between domain-specific depending on the measure used (Table 4), with the
cognitive marker variables and mood measures re- MMPI appearing most sensitive to depression and
vealed that STAIS scores were significantly correlated roughly equal to the STAIS in terms of sensitivity to
with TMTB scores (Table 3). No other significant as- symptoms of anxiety.
 Chemobrain in Breast Carcinoma?/Wefel et al. 471

Exploratory analyses were conducted to identify

demographic, disease-related, and treatment-related
factors that may be associated with the presence of
cognitive impairment (i.e., OCFI-I) before chemother-
apy. An adjusted ␣ level of 0.01 was chosen for the
assessment of statistical significance, given our desire
to correct for multiple comparisons and minimize the
possibility of Type I error. Association between OCFI
and age and education were explored using indepen-
dent-sample t tests. Association of disease-related and
treatment-related variables with OCFI were examined
using chi-square analyses. All disease- and treatment-
related variables were dichotomized. For disease
stage, women were divided into 2 groups: women with
Stage I or II disease (n ⫽ 74) and women with Stage
IIIA disease (n ⫽ 10). Women also were divided into 2
groups on the basis of extent of surgery: women who FIGURE 1. Overall Cognitive Function Index (OCFI) in relation to affective
underwent surgical resection (n ⫽ 41) and women status. CSD-I: clinically significant distress–impaired; CSD-NI: clinically signif-
who underwent biopsy only (n ⫽ 43). Women who icant distress–not impaired; OCFI-I: OCFI-impaired; OCFI-NI: OCFI–not im-
previously had received chest wall irradiation (n ⫽ 10) paired. Pearson ␹2(1, N ⫽ 78) ⫽ 9.90; P ⫽ 0.002.
were distinguished from women who had never re-
ceived irradiation (n ⫽ 74). Women who were post-
menopausal either naturally or due to surgery were
grouped together (n ⫽ 48) and were compared with
women who were either premenopausal or perimeno-
pausal (n ⫽ 36). Postmenopausal women with avail-
able data on HRT use were classified as HRT users (n
⫽ 30) if they had any history of HRT use or as HRT
nonusers (n ⫽ 18) if they had never used HRT.
Patients in the OCFI-I and OCFI-NI groups did not
differ significantly with regard to age (t[82] ⫽ 2.02; P
⫽ 0.05; mean ⫾ SD: OCFI-I, 53.1 ⫾ 9.8 years; OCFI-NI,
48.9 ⫾ 8.4 years) or education (t[82] ⫽ 1.92; P ⫽ 0.06;
mean ⫾ SD: OCFI-I, 13.2 ⫾ 2.7 years; OCFI-NI, 14.4
⫾ 2.4 years). Women who were postmenopausal (Fig.
1), had no history of HRT use (Fig. 2), or had under-
gone lumpectomy/mastectomy (Fig. 3) were nearly FIGURE 2. Overall Cognitive Function Index (OCFI) in relation to menopausal
status. Postmeno: postmenopausal; Premeno: premenopausal; OCFI-I: OCFI-
twice as likely to be cognitively impaired compared
impaired; OCFI-NI: OCFI–not impaired. Pearson ␹2(1, N ⫽ 84) ⫽ 4.22; P
with women who did not have such characteristics;
⫽ 0.04.
however, for none of these differences was P ⱕ 0.01.
Women with a classification of OCFI-I did not differ
significantly from women with a classification of evaluation of cognitive functioning against which
OCFI-NI in terms of any other demographic, disease- postchemotherapy changes in performance could be
related, or treatment-related variable. compared. A longitudinal design of this nature is nec-
essary to clearly demonstrate the relation between
DISCUSSION adjuvant chemotherapy and cognitive functioning.
Adjuvant chemotherapy has been reported to be as- In the current study, we found that 36% of women
sociated with the development of both early and de- with breast carcinoma demonstrated impaired overall
layed cognitive dysfunction in a population of women cognitive function before the initiation of adjuvant
with breast carcinoma. However, methodologic limi- chemotherapy. Cognitive impairment in the domains
tations in all studies on this topic that have been of verbal learning and memory was observed signifi-
published to date make these conclusions controver- cantly more frequently in the study cohort than in the
sial.35,36 Perhaps most significantly, previous studies comparison group of normal control patients. In ad-
failed to include a pretreatment neuropsychologic dition, differences in psychomotor processing speed
472 CANCER August 1, 2004 / Volume 101 / Number 3

treated with chemotherapy had performance scores

that fell within normal ranges makes attributions of
neurotoxicity in these patients particularly dubious.
Across these studies, the frequency of cognitive dys-
function ranged from 17% to 75%. Given the current
documentation of objective cognitive impairment be-
fore adjuvant systemic therapy, a large proportion of
patients in previously published reports who per-
formed at levels below what was expected when they
were assessed after chemotherapy may well have per-
formed at that same level before chemotherapy. Those
previous studies, therefore, may have overestimated
the true incidence of decreases in cognitive function-
ing secondary to chemotherapy.
The assessment of cognitive impairment in the
current study involved comparison of patients’ scores
on measures of cognitive skills with published norma-
FIGURE 3. Overall Cognitive Function Index (OCFI) in relation to hormone tive data. We propose that test performance scores for
replacement therapy (HRT) status. HRT nonusers had no history of HRT use, a group of individuals without breast carcinoma con-
whereas HRT users did have a history of HRT use. OCFI-I: OCFI-impaired; stituted a valid control for the purposes of the study.
OCFI-NI: OCFI–not impaired. Pearson ␹2(1, N ⫽ 48) ⫽ 2.29; P ⫽ 0.13. Women diagnosed with nonmetastatic breast carci-
noma are not commonly believed to experience neu-
rologic dysfunction. Thus, their performance is ex-
pected to be similar to that of individuals with similar
demographic backgrounds (e.g., in terms of age and
education). The use of normative data means that
study patients are being compared with just such a
control group. Normative data often have the advan-
tage of representing much larger populations than the
control populations recruited by individual investiga-
tors on a study-by-study basis. This leads to two im-
portant design features: 1) stratification according to
relevant demographic variables (e.g., age and educa-
tion) can be performed, so that the study group is
compared with appropriately matched control pa-
tients; and 2) the larger cohort associated with the
normative data provides a more stable estimate of the
mean and SD in the population from which the rep-
resentative sample was drawn, and thus a more accu-
FIGURE 4. Overall Cognitive Function Index (OCFI) in relation to extent of rate and valid estimate of the true score for the control
surgery. Lump/mast: lumpectomy or mastectomy; OCFI-I: OCFI-impaired; OCFI- group with which the study cohort is compared.
NI: OCFI–not impaired. Pearson ␹2(1, N ⫽ 84) ⫽ 4.95; P ⫽ 0.03. This approach has been criticized, because it fails
to control for potential differences in emotional dis-
tress that may accompany being diagnosed with a
and attention, nonverbal memory, confrontational medical illness. Thus, a similarly affected control
naming, complex visuoconstruction, and fine motor group often is desirable to control for this potentially
dexterity approached significance. In five previously confounding influence on cognitive function. How-
published cross-sectional studies that assessed pa- ever, because the current cohort was untreated, it was
tients’ cognitive status after chemotherapy, cognitive not possible to recruit another breast carcinoma co-
dysfunction was defined inconsistently and often far hort that would control for issues of emotional distress
less rigorously than in the current investigation, with while not also controlling for our variable of interest,
these definitions sometimes including performance cognitive function in women with breast carcinoma
scores that were within the ranges for normal control before treatment. The use of another medically af-
patients. The finding that many patients who were fected population also is problematic, because there is
 Chemobrain in Breast Carcinoma?/Wefel et al. 473

evidence that the presence of disease without direct diac and noncardiac surgery on patients’ postacute
CNS involvement may nonetheless have effects on the cognitive function, with some investigations demon-
CNS in some instances. For example, newly diagnosed strating declines in postoperative cognitive func-
(i.e., untreated) patients with small cell lung carci- tion41,42 and others reporting no significant decline
noma who had no evidence of CNS involvement ex- approximately 1 month after surgery.43 All women in
hibited cognitive dysfunction before the initiation of the current study who underwent lumpectomy or
any therapy,37 further underscoring the importance of mastectomy subsequently underwent neuropsycho-
obtaining pretreatment measures of cognitive func- logic evaluation an average of 53 days (SD, 39.8 days)
tion. Thus, the use of another population that is af- after surgery. Follow-up analyses in the current sam-
fected by disease may also obscure the very phenom- ple did not detect significant differences between pa-
enon that is of interest to the investigator. Due to tients who exhibited cognitive impairment (OCFI-I;
these considerations, we chose to use relevant norma- mean ⫾ SD, 63.8 ⫾ 50.3 days) and patients with a
tive data as our control and supplemented our assess- classification of cognitively not impaired (OCFI-NI;
ment with measures of emotional well-being to inves- mean ⫾ SD, 44.4 ⫾ 25.8 days) in terms of the length of
tigate the correlation between emotional distress and time between surgery and neuropsychologic evalua-
cognitive function. tion. Thus, there is no evidence that women who were
Methodologic limitations notwithstanding, a con- assessed more acutely postoperatively were more
sistent pattern of cognitive dysfunction attributable to likely to exhibit cognitive impairment.
chemotherapy has not emerged from previously pub- Alterations in women’s hormonal milieu report-
lished studies. The involved domains of cognitive edly have been associated with changes in cognitive
function have included learning and memory, pro- function. Although controversy remains regarding the
cessing speed, attention, and visuoperception. The cognitive effects of HRT,44 there is evidence that sur-
variability in findings of cognitive dysfunction may be gical or chemical induction of menopause is associ-
attributable in part to the choice of measures used in ated with impairments in cognitive function and
each study and the differential sensitivity of each mea- mood.45,46 The presence of affective distress was asso-
sure, as suggested in the current investigation. In gen- ciated with cognitive impairment in the current inves-
eral, researchers interested in studying perturbations tigation; however, it was not causally related to cog-
of the underlying neural networks that are responsible nitive impairment. Thus, depression, anxiety, and
for various cognitive functions may wish to employ cognitive dysfunction appear to be separable, albeit
the most sensitive measure for each specific cognitive comorbid, conditions in this population.
domain. The effects of a treatment (or any neurologic We recognize that a short screening measure that
condition), as demonstrated previously,38 may be ob- can be administered and interpreted by a variety of
servable only when a cognitive system is stressed suf- health care professionals with minimal training in
ficiently so that it unmasks the dysfunction and neuropsychology would be highly desirable. Unfortu-
overwhelms potential compensatory mechanisms. nately, the sensitivity of these measures (e.g., the
Relatively insensitive tests, such as cognitive screening Mini-Mental State Examination) in detecting cognitive
tools (e.g., the Mini-Mental State Examination39), tend disturbances that are not severe enough to be consid-
to underestimate the nature and extent of cognitive ered dementia is poor.40 Currently, clinicians should
dysfunction, often lack specificity for any particular continue discussing cognitive functioning with their
cognitive domain, and may fail to capture subtle but patients and should refer patients for neuropsycho-
meaningful changes in a patient’s cognitive status.40 logic evaluation if they or their family members report
We also attempted to discern factors related to the changes in their ability to think. In this way, a com-
occurrence of cognitive impairment in patients with prehensive evaluation may be undertaken that can
breast carcinoma. Although correlations between cog- yield differential diagnostic information and provide
nitive function before chemotherapy and demo- treatment recommendations.
graphic, disease-related, and clinical characteristics To our knowledge, the current report is the first to
did not exceed our adjusted ␣ value for determining document the presence of cognitive impairment in a
statistical significance, several provocative trends cohort of patients with breast carcinoma without CNS
emerged. Patients who underwent more invasive sur- involvement before the initiation of systemic adjuvant
gery (lumpectomy or mastectomy), patients who were therapy. Although the current study does not afford a
postmenopausal, and patients who had not previously comprehensive examination of the many potential eti-
used any HRT appeared to have a greater risk of pre- ologic agents responsible for this dysfunction, it is
senting with cognitive impairment. speculated that both host-related factors and disease-
Debate continues about the adverse effects of car- related factors may be involved. Research is actively
474 CANCER August 1, 2004 / Volume 101 / Number 3

pursuing host-related, or soil, characteristics, such as 7. Lipp HP. Neurotoxicity (including sensory toxicity) induced
genetic polymorphisms, immune reactivity, nutri- by cytostatics. In: Lipp HP, editor. Anticancer drug toxicity
prevention, management, and clinical pharmacokinetics.
tional factors, hormonal histories, or lack of cognitive
New York: Marcel Dekker, 1999:431– 454.
reserve, that are associated with an increased likeli- 8. Brezden CB, Phillips KA, Abdolell M, Bunston T, Tannock IF.
hood of developing cognitive dysfunction either be- Cognitive function in breast cancer patients receiving adju-
fore or in concurrence with the administration of ad- vant chemotherapy. J Clin Oncol. 2000;18:2695–2701.
juvant therapy. In addition, disease-related, or seed, 9. Schagen SB, van Dam FS, Muller MJ, et al. Cognitive deficits
after postoperative chemotherapy for breast carcinoma.
factors, including tumor gene mutations, induction of
Cancer. 1999;85:640 – 650.
proinflammatory cytokines, and paraneoplastic disor- 10. van Dam FS, Schagen SB, Muller MJ, et al. Impairment of
ders, as well as interactions between disease-related cognitive function in women receiving adjuvant treatment
factors and host-related factors, may contribute to the for high-risk breast cancer: high-dose versus standard-dose
development of cognitive dysfunction. Interdiscipli- chemotherapy. J Natl Cancer Inst. 1998;90:210 –218.
nary investigations of both the behavioral expression 11. Wieneke MH, Dienst ER. Neuropsychological assessment of
cognitive functioning following chemotherapy for breast
of perturbations within the neural networks responsi-
cancer. Psychooncology. 1995;4:61– 66.
ble for cognition and the potential mechanistic con- 12. Ahles TA, Saykin AJ, Furstenberg CT, et al. Neuropsycho-
tributors to the observed neurotoxicity will assist in logic impact of standard dose systemic chemotherapy in
the development of rational, targeted therapeutic op- long-term survivors of breast cancer and lymphoma. J Clin
tions (e.g., cytokine antagonists or neuroprotectants) Oncol. 2002;20:485– 493.
13. Tchen N, Juffs HG, Downie FP, et al. Cognitive function,
in the future.
fatigue, and menopausal symptoms in women receiving
There appears to be justification for concern re- adjuvant chemotherapy for breast cancer. J Clin Oncol.
garding the potential neurotoxicities associated with 2003;21:4175– 4183.
systemic antitumor agents used therapeutically to 14. Lezak MD. Neuropsychological assessment (3rd edition).
treat and cure breast carcinoma. However, our current New York: Oxford University Press, 1995.
understanding of the cognitive and neurobehavioral 15. Wechsler D. Wechsler Adult Intelligence Scale—revised. San
Antonio: The Psychological Corporation, 1981.
effects of these treatments is extremely limited. This 16. Wechsler D. Wechsler Adult Intelligence Scale (3rd edition).
issue is of significant clinical importance given the San Antonio: The Psychological Corporation, 1997.
prevalence of breast carcinoma; the increased use of 17. Wechsler D. Wechsler Memory Scale (3rd edition). San An-
chemotherapy as adjuvant therapy; the increasing use tonio: The Psychological Corporation, 1997.
of more aggressive dosing schedules; the increasing 18. Reitan RM, Davison LA. Clinical neuropsychology: current
status and applications. New York: Hemisphere Publishing
survival rates being achieved; and patients’ natural
Corp., 1974.
desire to return to their normal occupational, aca- 19. Reitan RM, Wolfson D. The Halstead–Reitan neuropsycho-
demic, and social pursuits. Thus, it is imperative that logical test battery. Tucson: Neuropsychology Press, 1985.
future investigations use well designed longitudinal 20. Brandt J. The Hopkins Verbal Learning Test: development of
methodologies that will assist in defining the relative a new memory test with six equivalent forms. Clin Neuro-
psychol. 1991;5:125–142.
risks and benefits inherent in the available treatments
21. Hannay HJ, Levin, HS. Selective Reminding Test: an exam-
so that truly informed treatment decisions can be ination of the equivalence of four forms. J Clin Exp Neuro-
made. The results of our own prospective longitudinal psychol. 1985;7:251–263.
trials will be available soon. 22. Fletcher JM. Memory for verbal and nonverbal stimuli in
learning disabled subgroups: analysis by selective remind-
REFERENCES ing. J Exp Child Psychol. 1985;40:244 –259.
1. Meyers CA. Issues of quality of life in neuro-oncology. In: 23. Osterrieth PA. Le test de copie d’une figure complexe. Arch
Vecht CJ, editor. Handbook of clinical neurology. Neuro- Psychologie. 1944;30:206 –356. [translated by Corwin J,
oncology, Part I. Brain tumors: principles of biology, diag- Bylsma FW]. Clin Neuropsychol. 1993;7:9 –15.
nosis and therapy. Volume 67. Amsterdam: Elsevier, 1997: 24. Benton AL, Hamsher K. Multilingual aphasia examination.
389 – 409. Iowa City: AJA Associates, 1983.
2. Meyers CA. Neurocognitive dysfunction in cancer patients. 25. Kaplan EF, Goodglass H, Weintraub S. The Boston Naming
Oncology. 2000;14:75–79. Test (2nd edition). Philadelphia: Lea & Febiger, 1983.
3. Silverberg E, Boring CC, Squires TS. Cancer statistics 1990. 26. DeFilippis NA, McCampbell E, Rogers P. Development of a
CA Cancer J Clin. 1990;40:9 –26. booklet form of the Category Test: normative and validity
4. Wolff AC. Systemic therapy. Curr Opin Oncol. 1999;11:468 – data. J Clin Neuropsychol. 1979;1:339 –342.
474. 27. Benton AL, Hamsher KD, Varney NR, Spreen O. Contribu-
5. Green M, Hortobagyi GN. Neoadjuvant chemotherapy for tions to neuropsychological assessment: a clinical manual.
operable breast cancer. Oncology (Huntingt). 2002;16:871– New York: Oxford University Press, 1983.
884. 28. Klove H. Clinical neuropsychology. Med Clin North Am.
6. Keime-Guibert F, Napolitano M, Delattre JY. Neurological 1963;46:1647–1658.
complications of radiotherapy and chemotherapy. J Neurol. 29. Beck AT, Steer RA. Beck Depression Inventory manual. San
1998;245:695–708. Antonio: The Psychological Corporation, 1987.
 Chemobrain in Breast Carcinoma?/Wefel et al. 475

30. Beck AT, Steer RA, Brown GK. BDI-II manual. San Antonio: 40. Meyers CA, Wefel JS. The use of the Mini-Mental State Exam-
The Psychological Corporation, 1996. ination to assess cognitive functioning in cancer trials: no ifs,
31. Hathaway SR, McKinley JC. Minnesota multiphasic person- ands, buts, or sensitivity. J Clin Oncol. 2003;21:3557–3558.
ality inventory. Minneapolis: National Computer Systems, 41. Newman MF, Kirchner JL, Phillips-Bute B, et al. Longitudi-
1970. nal assessment of neurocognitive function after coronary-
32. Spielberger CD, Gorsuch RL, Lushene R, Vagg PR, Jacobs artery bypass surgery. N Engl J Med. 2001;344:395– 402.
GA. The state-trait anxiety inventory. Palo Alto: Consulting 42. Rasmussen L, Johnson T, Kuipers HM, et al. Does anaes-
Psychologists Press, 1983. thesia cause postoperative cognitive dysfunction? A ran-
33. Beck AT, Steer RA. Beck Anxiety Inventory manual. San domized study of regional versus general anaesthesia in
Antonio: The Psychological Corporation, 1993. 438 elderly patients. Acta Anaesthesiol Scand. 2003;47:
34. Greene RL. The MMPI-2/MMPI: an interpretive manual. 260 –266.
Boston: Allyn and Bacon, 1991.
43. Browndyke JN, Moser DJ, Cohen RA, et al. Acute neuro-
35. Bender CM, Paraska KK, Sereika SM, Ryan CM, Berga SL.
psychological functioning following cardiosurgical inter-
Cognitive function and reproductive hormones in adjuvant
ventions associated with the production of intraoperative
therapy for breast cancer: a critical review. J Pain Symptom
cerebral microemboli. Clin Neuropsychol. 2002;16:463–
Manage. 2001;21:407– 424.
471.
36. Phillips KA, Bernhard J. Adjuvant breast cancer treatment
and cognitive function: current knowledge and research 44. Zec RF, Trivedi MA. The effects of estrogen replacement
directions. J Natl Cancer Inst. 2003;95:190 –197. therapy on neuropsychological functioning in postmeno-
37. Meyers CA, Byrne KS, Komaki R. Cognitive deficits in pa- pausal women with and without dementia: a critical and
tients with small cell lung cancer before and after chemo- theoretical review. Neuropsychol Rev. 2002;12:65–109.
therapy. Lung Cancer. 1995;12:231–235. 45. Sherwin B. Estrogen and cognitive functioning in women.
38. Korol DL. Enhancing cognitive function across the life span. Proc Soc Exp Biol Med. 1998;217:17–22.
Ann N Y Acad Sci. 2002;959:167–179. 46. Varney NR, Syrop C, Kubu CS, Struchen M, Hahn S, Franzen
39. Folstein MF, Folstein SE, McHugh PR. Mini-Mental State: a K. Neuropsychologic dysfunction in women following leu-
practical method for grading the state of patients for the prolide acetate induction of hypoestrogenism. J Assist Re-
clinician. J Psychiatr Res. 1975;12:89 –98. prod Genet. 1993;10:53–57.