Leukemia & Lymphoma, 2014; Early Online: 1–6 © 2014 Informa UK, Ltd. ISSN: 1042-8194 print / 1029-2403 online DOI: 10.3109/10428194.2014.915546
ORIGINAL ARTICLE: CLINICAL
Leuk Lymphoma Downloaded from informahealthcare.com by Monash University on 12/04/14 For personal use only.
Post-chemotherapy cognitive impairment in patients with B-cell non-Hodgkin lymphoma: a first comprehensive approach to determine cognitive impairments after treatment with rituximab, cyclophosphamide, doxorubicin, vincristine and prednisone or rituximab and bendamustine Philipp Zimmer1, Andreas Mierau2, Wilhelm Bloch1, Heiko K. Strüder2, Thorben Hülsdünker2, Alexander Schenk1, Leonie Fiebig1, Freek T. Baumann1, Moritz Hahn3, Nina Reinart4, Michael Hallek4 & Thomas Elter4 1Department for Molecular and Cellular Sports Medicine and 2Institute of Movement and Neurosciences, German Sport
University Cologne, Germany, 3Department for Medical Statistics, Informatics and Epidemiology, University of Cologne, Germany and 4Department I of Internal Medicine, Center for Integrated Oncology Koeln Bonn, Center of Excellence on “Cellular Stress Responses in Aging-Associated Diseases”, University of Cologne, Germany caused by chemotherapeutic agents, patients suffering from other psychological disorders (i.e. depression) also report a decline of their cognitive abilities [2]. In view of the prolonged survival rates which most patients with cancer experience through advances in medical treatment, it is essential to learn more about treatment-specific side effects. To date, the mechanisms for PCCI are still under investigation and not well understood. Valid and objective methods to assess specific, PCCI associated side effects are not yet established [3]. In a meta-analysis, Hodgson et al. [4] reported that verbal memory and executive functions are commonly affected by chemotherapies. However, Hutchinson and colleagues [5] found that chemotherapy-induced subjective cognitive impairments correlated with objective impairments in only eight of 24 studies. This variability in the occurrence of PCCI as well as the weak correlation of objective and self-reported parameters might be due to a sampling error, variability of time intervals after chemotherapy (immediately after versus 10 years after therapy), different assessments or a focus on specific tumors, such as breast cancer [6]. Besides subjective and objective cognitive tests, functional neuroimaging techniques might have the potential to provide new insights into the underlying mechanisms of PCCI. Scherling and Smith [3] described numerous studies using different magnetic resonance imaging (MRI) and positron emission tomography (PET) techniques to assess PCCI. In contrast to the above-mentioned techniques, only a few studies have implemented electroencephalographic (EEG) analysis to investigate the influence of cancer therapies on brain activity with regard to PCCI. Only Schagen
Abstract To assess the effects of chemoimmunotherapy on postchemotherapy cognitive impairments (PCCI) in patients with B-cell non-Hodgkin lymphoma (NHL), we used objective and subjective measures of cognitive functions in combination with serum parameters and neuroelectric recordings. Self-perceived status of cognition, fatigue and emotional functioning were reduced in patients (n ⫽ 30) compared to healthy controls (n ⫽ 10). Cognitive performance was impaired in patients with NHL compared to controls and a norm sample (n ⫽ 1179). PCCI was more severe in patients treated with rituximab and bendamustine (BR) than in patients who received R in combination with CHOP (cyclophosphamide, doxorubicin, vincristine, prednisone) polychemotherapy (R-CHOP). Individual alpha peak frequency and serum brain-derived neurotrophic factor (BDNF) levels in patients with NHL correlated with accuracy in the objective cognition test. Higher serum interleukin-6 (IL-6) concentrations were associated with higher fatigue levels. Patients with NHL and especially those who were treated with BR were affected by PCCI. BDNF and IL-6 might be involved in the pathogenesis of PCCI and fatigue. Keywords: Chemobrain, lymphoma, EEG, fatigue, cognition, cytokines
Introduction Post-chemotherapy cognitive impairment (PCCI), also known as “chemobrain,” is a well-known side effect of cancer therapies, affecting 17–75% of all patients with cancer [1]. Although PCCI focuses only on cognitive impairments
Correspondence: Philipp Zimmer, German Sport University Cologne, Department for Molecular and Cellular Sports Medicine, Am Sportpark Müngersdorf 6, 50933 Köln, Germany. Tel: 0049(0)221-4982-4821. Fax: 0049(0)221-4982-8370. E-mail: [email protected] Received 11 March 2014; revised 7 April 2014; accepted 12 April 2014
1
Leuk Lymphoma Downloaded from informahealthcare.com by Monash University on 12/04/14 For personal use only.
2
P. Zimmer et al.
et al. [7] and Kreukels et al. [8] combined EEG analyses with objective cognitive tests in patients with breast cancer to evaluate cognitive deficits. Their results indicated that the latency and amplitude of event related potentials are influenced by chemotherapy. These few studies have similar limitations to the studies discussed above. The infrequent application of EEG in the context of PCCI is remarkable, since Angelakis et al. [9] and Richard et al. [10] described a correlation between the individual alpha peak frequency (iAPF) and cognitive functions. The alpha rhythm is the dominant EEG pattern in the human scalp. iAPF is defined as the peak frequency within the range of the alpha frequency (7.5–12.5 Hz). For an extensive review see the article by Klimesch [11]. Rodin and Ahles [6] also made the criticism that objective and subjective cognition data are rarely combined with serum parameters when investigating PCCI. Although knowledge about the function of pro-inflammatory cytokines (i.e. interleukin-6 [IL-6] and tumor necrosis factor-α [TNF-α]) related to cognitive impairment is rudimental, they might be involved in generating and maintaining side effects such as fatigue and PCCI [12,13]. Besides pro-inflammatory cytokines, altered concentrations of neurotrophic factors such as brain-derived neurotophic factor (BDNF) may contribute to changes in cognitive function. BDNF is known to promote the growth and differentiation of neuronal stem cells and to support the survival of existing neurons. In addition, elevated BDNF levels are associated with enhanced long-term potentiation [14]. The aim of the present clinical trial was to determine the influence of chemoimmunotherapy in patients with non-Hodgkin lymphoma (NHL) on objective and subjective cognitive functions and potential (neuro-) physiological correlates. For this purpose, subjective and objective cognitive assessment tools were combined with EEG analyses and potential PCCI-relevant cytokine measurements in a group of patients with NHL and a healthy control group. In a subgroup analysis, patients who had been treated for indolent NHL with the monoclonal antibody rituximab and bendamustine (BR) were separated from those with aggressive NHL who had received R in combination with a poly-chemotherapy regimen, cyclophosphamide, doxorubicin, vincristine and prednisone (R-CHOP).
Participants and methods The study protocol was reviewed and approved by the local ethics committee in accordance with the Declaration of Helsinki. A trial registration was conducted at the German Clinical Trial Register, which is part of the world Health Organization (WHO) International Clinical Trials Registry Platform (No. DRKS0004743). All participants provide written informed consent prior to participation. Thirty-two patients were invited to participate. Two patients refused, leading to a recruitment rate of 93.75%.
Participants Over a period of 18 months, 30 patients with B-cell NHL (mean age 63.7 years, standard deviation [SD] ⫽ 12.8, 11
female, 19 male) and 10 healthy controls (mean age 56.6 years, SD ⫽ 10.7, three female, seven male) were recruited between March 2012 and September 2013. Of the 30 patients who received first-line therapy, 16 (five female, 11 male) were treated with BR whereas the other 14 patients (three female, 11 male) received R-CHOP. All patients were assessed within the first 3 months after their chemotherapy ended. The following inclusion criteria were used: first-line therapy with BR or R-CHOP had to be completed between 4 and 12 weeks prior to inclusion in the trial, and patients had to be older than 18 years. Patients who underwent second-line chemotherapy or radiation therapy or presented any condition (e.g. orthopedic restriction) preventing participation in the cognition tests were excluded.
Workflow Participants were examined once, between 9 a.m. and 10 a.m. After serum samples were taken, patients completed the European Organisation for Research and Treatment of Cancer Quality of Life Questionnaire-C30 (EORTC QLQ-C30), followed by a 5 min resting state EEG with closed eyes and a computer based cognition test.
Objective cognition assessment Executive function and attention were chosen as objective cognition parameters because these abilities have been described to be influenced by chemotherapy [4]. To determine the named abilities, we used the computer based Wiener test system (Schuhfried, Vienna, Austria). During the determination test (DT), patients had to react to visual and optical stimuli by pressing colored buttons on a keyboard and foot pedals for both feet. The test started after completing a standardized test instruction, including a practice period in order to ensure that the tasks were fully understood. All stimuli are presented adaptively. That is, the speed of presentation is adapted to the ability level of the participant. In detail, the DT can measure accuracy and processing speed under time pressure. According to the supervisory attentional system theory [15], the DT is an appropriate assessment to measure executive function. The main outcome is “correct responses.” This variable describes the individual processing limit under time pressure. Additional outcomes are “incorrect responses” and “omitted responses.” The internal consistency reliability for the variables ranges from r ⫽ 0.98 to r ⫽ 0.99. The results are presented as raw scores as well as percent rank compared to an age- and grade of education-matched norm sample, including a total number of 1179 healthy subjects. As described by the manufacturer, the majority of the norm sample has percent ranks between 25 and 75. Thus, percent ranks lower than 25 were considered as a weak performance, whereas percent ranks above 75 were considered as a strong performance [16].
Self-perceived status of cognition, fatigue and emotional functioning Subjective cognition status, fatigue and emotional functioning were evaluated using the subscales for cognitive function (CF), emotional function (EF) and fatigue (FA) of the EORTC
Chemobrain after B-cell non-Hodgkin lymphoma QLQ-C30 [17]. We decided to use this questionnaire because it includes all demanded domains. Additionally, this established questionnaire is widely used, making our data comparable to the outcomes of other studies.
Leuk Lymphoma Downloaded from informahealthcare.com by Monash University on 12/04/14 For personal use only.
EEG recording and data analysis Electroencephalography was recorded from 15 scalp locations using an electrode cap and a 16-channel EEG amplifier (Brain Products, Munich, Germany). Recording positions were equally distributed over the scalp (Fp1, Fp2, T7, F3, Fz, F4, T8, C3, Cz, C4, P3, Pz, P4, O1, O2) according to the international 10:20 system, with the electrical reference located on position FCz. Additionally, one electrooculographic electrode was used to detect vertical and horizontal eye movements and related artifacts in the EEG signal. Electrode impedances were kept below 5 kΩ. The sampling rate was set to 1000 Hz. For data analysis, Brain Vision Analyzer 2 software was used (Brain Products, Munich, Germany). An infinite-impulse response (IIR) band-pass filter (2–30 Hz, 48 dB/oct) was applied to the data to exclude low frequency artifacts as well as frequencies that were not relevant for the purpose of the study. In a second step, EEG data were re-referenced to the average potential of the 15 recording electrodes. Resting state data were then subdivided into segments of 4 s and baseline corrected. An independent component analysis (ICA) algorithm optimized for ocular artifact correction was applied, and those components representing ocular artifacts were excluded from further analysis. Subsequently, a semi-automatic artifact rejection (maximum allowed voltage step: 150 μV; maximal allowed voltage difference within 200 ms: 100 μV) was used to reject residual artifacts. Spectral power was calculated using fast Fourier transformation (FFT) (frequency resolution: 0.244 Hz) and frequency spectra were averaged across all segments. The iAPF was defined as the frequency indicating the highest power value within the 8–13 Hz range [18].
Serum parameter measurement IL-6 and BDNF concentrations of serum samples were analyzed using enzyme-linked immunosorbent assays
according to the manufacturer’s protocol (R&D systems, Minneapolis, MN).
Statistical analysis Statistical analysis was performed using SPSS 20 (IBM, Rochester, MN). In a first step, comparisons between the patient group and the control group were conducted with an unpaired t-test. Variance homogeneity was confirmed using the Levene test. In a second step, Spearman correlations were calculated for all variables. For subgroup analysis, analysis of variance (ANOVA) with subsequent Bonferroni post hoc test was performed. Furthermore, Spearman correlations were calculated for the subgroups of patients. To compare participants’ data for the objective cognition test with regard to those of the norm sample, median percent ranks were analyzed by Mann–Whitney U-test. The significance level was set at α ⫽ 0.05.
Results Participant characteristics In terms of age, no group differences could be detected (p ⫽ 0.123). According to the disease incidence rate, sex distribution was not equal among the groups (see “Participants and methods” above).
Objective cognition assessment outcomes The t-test revealed significant group differences for the main outcome of the objective cognition test (“correct responses”). The number of “correct responses” was significantly lower in patients when compared to the control group (p ⫽ 0.003). “Incorrect responses” and “omitted responses” did not differ significantly between the groups (Table I). Patients who were treated with BR achieved significantly lower numbers of “correct responses” than the control group (p ⬍ 0.001) and the R-CHOP group (p ⫽ 0.005). Additionally, patients who received BR had significantly more “omitted responses” than the control group (p ⫽ 0.034). The outcomes of the R-CHOP group did not differ from those of the healthy controls (Table II).
Table I. Results of all outcomes of patient and control groups†. Control group Objective cognition test outcomes Correct responses 228 (7.35) Incorrect responses 13.5 (4.10) Omitted responses 12.2 (2.42) Subjective perception parameters Cognition 88.3 (4.34) Fatigue 9.96 (4.20) Emotional function 81.5 (6.03) EEG data iAPF frontal region 9.26 (0.30) iAPF central region 9.41 (0.28) iAPF parietal region 9.47 (0.25) Cytokines IL-6 (pg/mL) 1,94 (0.19) 15.40 (3.01) BDNF (μg/mL)
3
Patients
p-Value
95% CI
186 (7.05) 16.1 (2.11) 18.4 (1.76)
0.003* 0.412 0.060
[⫺ 67.8; ⫺ 15.3] [⫺ 6.1; 11.4] [⫺ 0.27; 12.6]
64.4 (5.04) 54.4 (4.70) 58.5 (5.26)
0.013* 0.000* 0.024*
[⫺ 42.4; ⫺ 5.40] [31.64; 57.33] [⫺ 42.9; ⫺ 3.16]
9.66 (0.09) 9.44 (0.09) 9.99 (0.15)
0.193 0.844 0.093
[0.23; 1.01] [⫺ 0.61; 0.68] [⫺ 0.88; 1.13]
3,67 (0.33) 14.23 (1.20)
0.000* 0.723
[0.95; 2.49] [⫺ 81.8; 58.4]
EEG, electroencephalographic; iAPF, individual alpha peak frequency; IL-6, interleukin-6; BDNF, brain-derived neurotrophic factor. *Significant difference, t-test. †Mean, standard error of the mean and 95% confidence interval (CI) are presented for all outcomes.
4
P. Zimmer et al. Table II. Results of all outcomes of subgroup analysis. ANOVA
Leuk Lymphoma Downloaded from informahealthcare.com by Monash University on 12/04/14 For personal use only.
Objective cognition test outcomes Correct responses ⬎ 0.001* Incorrect responses 0.837 Omitted responses 0.027* Subjective perception parameters Cognition 0.006* Fatigue ⬍ 0.001* Emotional function 0.078 EEG data iAPF frontal region 0.040* iAPF central region 0.704 iAPF parietal region 0.185 Cytokines IL-6 (pg/mL) 0.020* 0.258 BDNF (μg/mL)
Controls vs. R-CHOP 0.354 1.00 1.00
Controls vs. BR 0.000* 1.00 0.034*
R-CHOP vs. BR 0.005* 1.00 0.167
0.005* ⬍ 0.001* 0.108
0.359 ⬍ 0.001* 0.165
0.137 1.00 1.00
1.00 1.00 0.757
0.051 1.00 0.204
0.219 1.00 1.00
0.051 1.00
0.028* 0.852
1.00 1.00
EEG, electroencephalographic; iAPF, individual alpha peak frequency; IL-6, interleukin-6; BDNF, brain-derived neurotrophic factor; ANOVA, analysis of variance; R-CHOP, rituximab, cyclophosphamide, doxorubicin, vincristine, prednisone; BR, rituximab, bendamustine. *Significant difference, p-value, ANOVA and Bonferroni post hoc.
The median percent rank of the outcomes “correct responses” and “omitted responses” showed that patients who were treated with BR had a median percent rank lower than 25. Patients of the R-CHOP group as well as the control group revealed a median percent rank within the usual range (25–75) of the norm sample (Figure 1). Subjects of the control group showed a significantly higher median percent rank regarding “correct responses” than the patient group and both patient subgroups. No group differences could be detected for “incorrect responses.” Median percent rank for “omitted responses” differed significantly between controls and patients as well as the BR subgroup (Figure 1).
Self-perceived outcomes Patients demonstrated significantly lower scores of subjective perception of cognition, fatigue and emotional functioning (Table I). Within the subgroup analysis, patients in the R-CHOP group reported significantly more cognitive impairments than the control group (p ⫽ 0.005), whereas no difference was found between the other groups. Fatigue data did not differ between the patient groups but were significantly elevated in comparison to the control group (Table II).
EEG data iAPF data did not differ between patients and controls. ANOVA of the frontal iAPF yielded a significant group effect (p ⫽ 0.040). The subsequent post hoc test showed a trend toward lower frontal iAPF values in the BR group compared to controls. However, this difference did not reach statistical significance (p ⫽ 0.051) (Table II).
Cytokine measurements IL-6 serum levels were significantly elevated in the patient group (p ⬍ 0.001) (Table I). The BR group exhibited significantly higher IL-6 serum concentrations than the healthy control group. The R-CHOP group also showed elevated IL-6 serum concentrations. However, the difference from the control group did not reach significance (p ⫽ 0.051) (Table II). BDNF levels did not show any group differences.
Correlations All significant correlations are listed in Table III. IL-6 serum concentrations correlated with fatigue levels (p ⫽ 0.003, r ⫽ 0.522) and age (p ⫽ 0.026, r ⫽ 0.406), but did not show any correlation with outcomes of the objective cognition test. Patients with higher BDNF levels had better results in the main variable (“correct responses”) of the DT (p ⫽ 0.041,
Figure 1. Boxplots of median percent rank for main variables of the objective cognition test for all groups and subgroups. Significant differences of the median percent ranks are presented with p-values. Percent ranks ranging from 25 to 75 are regarded as normal (gray shading represents data of the norm sample). Percent rank smaller than 25 is seen as a weak development of the variable. In addition to medians, means are indicated by ⫹.
Chemobrain after B-cell non-Hodgkin lymphoma
5
Leuk Lymphoma Downloaded from informahealthcare.com by Monash University on 12/04/14 For personal use only.
Table III. Significant correlations of all outcomes of patient group and both subgroups*. Patient group (n ⫽ 30) IL-6 BDNF iAPF parietal region iAPF frontal region Age Age Age BR (n ⫽ 16) IL-6 iAPF central region iAPF central region iAPF parietal region iAPF frontal region iAPF frontal region Age Age R-CHOP (n ⫽ 14) iAPF frontal region iAPF parietal region
Outcome
p-Value, correlation coefficient
Fatigue Correct responses Omitted responses Emotional functioning Correct responses Subjective perception of cognition IL-6
p ⫽ 0.003, r ⫽ 0.522 p ⫽ 0.041, r ⫽ 0.375 p ⫽ 0.017, r ⫽ ⫺ 0.432 p ⫽ 0.027, r ⫽ ⫺ 0.403 p ⫽ 0.008, r ⫽ ⫺ 0.474 p ⫽ 0.001, r ⫽ 0.586 p ⫽ 0.029, r ⫽ 0.428
Fatigue Correct responses Incorrect responses Omitted responses Emotional functioning Subjective perception of cognition Subjective perception of cognition IL-6
p ⫽ 0.034, r ⫽ 0.532 p ⫽ 0.033, r ⫽ 0.534 p ⫽ 0.025, r ⫽ ⫺ 0.558 p ⬍ 0.001, r ⫽ ⫺ 0.773 p ⫽ 0.024, r ⫽ ⫺ 0.559 p ⫽ 0.047, r ⫽ ⫺ 0.503 p ⫽ 0.018, r ⫽ 0.581 p ⫽ 0.039, r ⫽ 0.601
Incorrect responses BDNF
p ⫽ 0.023, r ⫽ 0.601 p ⫽ 0.010, r ⫽ ⫺ 0.633
EEG, electroencephalographic; IL-6, interleukin-6; BDNF, brain-derived neurotrophic factor; iAPF, individual alpha peak frequency; BR, rituximab, bendamustine; R-CHOP, rituximab, cyclophosphamide, doxorubicin, vincristine, prednisone. *All significant correlations between objective cognition test, subjective perception, EEG data and serum parameters of patient group and subgroups.
r ⫽ 0.375). We found correlations between the iAPF of the parietal region and “omitted responses” (p ⫽ 0.017, r ⫽ ⫺ 0.432) and the iAPF of the frontal cortex and emotional functioning (p ⫽ 0.027, r ⫽ 0.403) in the DT. A similar pattern of correlation was obtained in the BR group (with the exception of the BDNF data). Additionally, in the BR group, significant correlations were found between all outcomes of the objective cognition test and the iAPF data. In contrast to the BR group, the described correlations were not detectable in patients who received R-CHOP.
Discussion This study is the first comprehensive analysis of the impact of chemoimmunotherapy on cognitive functions of patients with B-cell NHL combined with serum and neurophysiological parameters. Furthermore, this is the first study comparing two different chemotherapy strategies (BR vs. R-CHOP) regarding their impact on cognitive abilities and associated physiological outcomes in patients with NHL. Our results indicate that patients with NHL have cognitive impairments within 3 months after the end of their chemoimmunotherapy. These results are in accordance with the patients’ subjective perception of cognition and with comparable studies [6,19]. Subgroup analyses revealed that patients treated with BR showed significantly lower performance scores in the objective cognition test than controls, the R-CHOP group and the age-matched norm sample. Indeed, one case report indicates that the application of BR is associated with a cognitive decline [20]. However, we expected that patients treated with R-CHOP, a regimen including one neurotoxic agent (vincristine) and one blood–brain barriercrossing drug (prednisone), would have elevated cognitive impairments. In contrast to Tholstrup et al. [21], we found an impact on the subjective perception of cognition after chemoimmunotherapy with R-CHOP. These findings were
not confirmed by the outcomes of the objective cognition assessment. Differences between controls and patients with NHL regarding the iAPF were not detected. However, a trend toward lower iAPF levels in the BR group was observed. Significant correlations of the iAPF and objective cognition outcomes were found within patients, and especially within the BR subgroup. These results confirmed other studies reporting a correlation of high iAPF values and selective aspects of cognition [9,10]. The detected correlation of frontal iAPF values and emotional functioning is in accordance with findings in depressive patients [22]. Therefore, regional iAPF measurements might be used for the objective evaluation of both cognitive impairments and emotional disorders. BDNF serum levels correlated significantly with the main outcome (“correct responses”) of the cognition test. These results confirm those of Oral et al. [23], who found reduced BDNF concentrations in depressive and cognitively impaired patients. We assume that serum BDNF levels may be used as an additional surrogate parameter to supplement the results of objective cognitive assessments. Additionally, they could help to understand the underlying mechanism of PCCI. The correlation of IL-6 serum levels and fatigue is in agreement with other studies, leading to the hypothesis that inflammation is a main cause of fatigue, independent of cancer type and therapy [12,13,24]. Elevated inflammation markers were also detected in various patient groups with depressive disorders, which would support the connection with fatigue and depression [25]. Further evidence for this relationship between inflammation and fatigue comes from animal studies [26]. However, the translational character of such findings is restricted. Future research in this field should focus on inflammation markers such as IL-6 within the brain and their origin. The use of IL-6 serum level as a single marker for these symptoms might not be sufficient. Inconspicuous IL-6 concentrations in patients in the R-CHOP group
Leuk Lymphoma Downloaded from informahealthcare.com by Monash University on 12/04/14 For personal use only.
6
P. Zimmer et al.
were not surprising, since the therapy protocol included the anti-inflammatory drug prednisone. The results of the present study should be considered within the context of its limitations. First, patients were not tested in a longitudinal design. Since numerous studies have shown cognitive impairments prior to chemotherapy, the phenomenon of PCCI might be biased by typical pretreatment symptoms such as depression, anxiety and stress [27,28]. Regarding the self-perceived status of cognition, fatigue and emotional functioning, we decided to use the well-established and often used EORTC QLQ-C30. The named domains of this questionnaire constitute only a few questions. One could use more detailed questionnaires; however, it should be taken into account that study subjects, and especially those with cognitive impairments, may have problems with completing a vast number of questionnaires. As objective assessments, other cognitive abilities such as memory should be included in future studies. Finally, studies could use a wider range of biomarkers in larger patient groups. In conclusion, our results reveal that patients with NHL, in particular when treated with BR, show cognitive impairments within 3 months after their chemoimmunotherapy. Objective cognitive assessments are necessary to detect PCCI. iAPF, as a neuroelectric index of cognition, as well as BDNF and IL-6 serum levels are biomarkers that deserve further study in the context of PCCI and fatigue. EEG analysis as well as BDNF levels might have the potential to support results from cognition tests and enable us to learn more about the underlying mechanisms.
Acknowledgements The authors would like to thank Dr. Pauline Schuhmacher and Mrs. Muriel Freudenberger for editorial support. Dr. Schuhmacher is supported by the German Federal Ministry of Research and Education (BMBF grant 01KN1106). Mrs. Freudenberger is supported by the Cologne Excellence Cluster for Cellular Stress Responses in AgingAssociated Diseases (CECAD) (DFG EXC 229). Furthermore, the authors would like to thank Mrs. Anke Schmitz and Mrs. Anika Voss for their technical support. Potential conflict of interest: Disclosure forms provided by the authors are available with the full text of this article at www.informahealthcare.com/lal.
References [1] Myers JS. Chemotherapy-related cognitive impairment: neuroimaging, neuropsychological testing, and the neuropsychologist. Clin J Oncol Nurs 2009;13:413–21. [2] van Dam FS, Schagen SB, Muller MJ, et al. Impairment of cognitive function in women receiving adjuvant treatment for high-risk breast cancer: high-dose versus standard-dose chemotherapy. J Natl Cancer Inst 1998;90:210–218. [3] Scherling CS, Smith A . Opening up the window into “chemobrain”: a neuroimaging review. Sensors (Basel) 2013;13:3169–3203. [4] Hodgson KD, Hutchinson AD, Wilson CJ, et al. A meta-analysis of the effects of chemotherapy on cognition in patients with cancer. Cancer Treat Rev 2013;39:297–304.
[5] Hutchinson AD, Hosking JR, Kichenadasse G, et al. Objective and subjective cognitive impairment following chemotherapy for cancer: a systematic review. Cancer Treat Rev 2012;38:926–934. [6] Rodin G, Ahles TA . Accumulating evidence for the effect of chemotherapy on cognition. J Clin Oncol 2012;30:3568–3569. [7] Schagen SB, Hamburger HL, Muller MJ, et al. Neurophysiological evaluation of late effects of adjuvant high-dose chemotherapy on cognitive function. J Neurooncol 2001;51:159–165. [8] Kreukels BP, Schagen SB, Ridderinkhof KR, et al. Electrophysiological correlates of information processing in breast-cancer patients treated with adjuvant chemotherapy. Breast Cancer Res Treat 2005;94:53–61. [9] Angelakis E, Lubar JF, Stathopoulou S. Electroencephalographic peak alpha frequency correlates of cognitive traits. Neurosci Lett 2004;16;371:60–63. [10] Richard Clark C, Veltmeyer MD, Hamilton RJ, et al. Spontaneous alpha peak frequency predicts working memory performance across the age span. Int J Psychophysiol 2004;53:1–9. [11] Klimesch W. EEG alpha and theta oscillations reflect cognitive and memory performance: a review and analysis. Brain Res Brain Res Rev 1999;29:169–195. [12] Seruga B, Zhang H, Bernstein LJ, et al. Cytokines and their relationship to the symptoms and outcome of cancer. Nat Rev Cancer 2008;8:887–899. [13] Ganz PA , Bower JE. Cancer related fatigue: a focus on breast cancer and Hodgkin’s disease survivors. Acta Oncol 2007;46: 474–479. [14] Huang EJ, Reichardt LF. Neurotrophins: roles in neuronal development and function. Annu Rev Neurosci 2001;24:677–736. [15] Norman DA , Shallice T. Attention to action: willed and automatic control of behaviour. In: Shapiro DL, Schwartz G, editors. Consciousness and self-regulation: advances in research. New York: Plenum Press ; 1986. pp 1–14. [16] Neuwirth W, Benesch M. Manual determination test. 35 ed. Mödling, Austria: Schuhfried GmbH; 2012. [17] Aaronson NK , Ahmedzai S, Bergman B, et al. The European Organization for Research and Treatment of Cancer QLQ-C30: a quality-of-life instrument for use in international clinical trials in oncology. J Natl Cancer Inst 1993;85:365–376. [18] Doppelmayr M, Kliemesch W, Pachinger T, et al. Individual differences in brain dynamics: Important implications for the calculation of event related band power. Biol Cybern 1998;79:49–57. [19] Jensen RE, Arora NK, Bellizzi KM, et al. Health-related quality of life among survivors of aggressive non-Hodgkin lymphoma. Cancer 2013;119:672–680. [20] Küllmer A, Herrmann G, Riecken B. [Rapid-progressive decrease of cognitive and physical functions in a B-cell non-Hodgkin’s lymphoma patient treated with rituximab/bendamustine]. Dtsch Med Wochenschr 2013;138:1355–1359. [21] Tholstrup D, Brown Pde N, Jurlander J, et al. Quality of life in patients with diffuse large B-cell lymphoma treated with dosedense chemotherapy is only affected temporarily. Leuk Lymphoma 2011;52:400–408. [22] Arns M, Drinkenburg WH, Fitzgerald PB, et al. Neurophysiological predictors of non-response to rTMS in depression. Brain Stimul 2012;5:569–576. [23] Oral E, Canpolat S, Yildirim S, et al. Cognitive functions and serum levels of brain-derived neurotrophic factor in patients with major depressive disorder. Brain Res Bull 2012;88:454–459. [24] Bower JE, Ganz PA , Irwin MR, et al. Inflammation and behavioral symptoms after breast cancer treatment: do fatigue, depression, and sleep disturbance share a common underlying mechanism? J Clin Oncol 2011;29:3517–3522. [25] Elderkin-Thompson V, Mintz J, Haroon E, et al. Executive dysfunction and memory in older patients with major and minor depression. Arch Clin Neuropsychol 2007;22:261–270. [26] Patki G, Solanki N, Atrooz F, et al. Depression, anxiety-like behavior and memory impairment are associated with increased oxidative stress and inflammation in a rat model of social stress. Brain Res 2013;1529:73–86. [27] Wefel JS, Lenzi R, Theriault RL, et al. The cognitive sequelae of standard-dose adjuvant chemotherapy in woman with breast carcinoma: results of a prospective, randomized, longitudinal trial. Cancer 2004;100:2292–2299. [28] Jansen CE, Cooper BA , Dodd MJ, et al. A prospective longitudinal study of chemotherapy-induced cognitive changes in breast cancer patients. Support Care Cancer 2011;10:1647–1656.
ORIGINAL ARTICLE: CLINICAL
Leuk Lymphoma Downloaded from informahealthcare.com by Monash University on 12/04/14 For personal use only.
Post-chemotherapy cognitive impairment in patients with B-cell non-Hodgkin lymphoma: a first comprehensive approach to determine cognitive impairments after treatment with rituximab, cyclophosphamide, doxorubicin, vincristine and prednisone or rituximab and bendamustine Philipp Zimmer1, Andreas Mierau2, Wilhelm Bloch1, Heiko K. Strüder2, Thorben Hülsdünker2, Alexander Schenk1, Leonie Fiebig1, Freek T. Baumann1, Moritz Hahn3, Nina Reinart4, Michael Hallek4 & Thomas Elter4 1Department for Molecular and Cellular Sports Medicine and 2Institute of Movement and Neurosciences, German Sport
University Cologne, Germany, 3Department for Medical Statistics, Informatics and Epidemiology, University of Cologne, Germany and 4Department I of Internal Medicine, Center for Integrated Oncology Koeln Bonn, Center of Excellence on “Cellular Stress Responses in Aging-Associated Diseases”, University of Cologne, Germany caused by chemotherapeutic agents, patients suffering from other psychological disorders (i.e. depression) also report a decline of their cognitive abilities [2]. In view of the prolonged survival rates which most patients with cancer experience through advances in medical treatment, it is essential to learn more about treatment-specific side effects. To date, the mechanisms for PCCI are still under investigation and not well understood. Valid and objective methods to assess specific, PCCI associated side effects are not yet established [3]. In a meta-analysis, Hodgson et al. [4] reported that verbal memory and executive functions are commonly affected by chemotherapies. However, Hutchinson and colleagues [5] found that chemotherapy-induced subjective cognitive impairments correlated with objective impairments in only eight of 24 studies. This variability in the occurrence of PCCI as well as the weak correlation of objective and self-reported parameters might be due to a sampling error, variability of time intervals after chemotherapy (immediately after versus 10 years after therapy), different assessments or a focus on specific tumors, such as breast cancer [6]. Besides subjective and objective cognitive tests, functional neuroimaging techniques might have the potential to provide new insights into the underlying mechanisms of PCCI. Scherling and Smith [3] described numerous studies using different magnetic resonance imaging (MRI) and positron emission tomography (PET) techniques to assess PCCI. In contrast to the above-mentioned techniques, only a few studies have implemented electroencephalographic (EEG) analysis to investigate the influence of cancer therapies on brain activity with regard to PCCI. Only Schagen
Abstract To assess the effects of chemoimmunotherapy on postchemotherapy cognitive impairments (PCCI) in patients with B-cell non-Hodgkin lymphoma (NHL), we used objective and subjective measures of cognitive functions in combination with serum parameters and neuroelectric recordings. Self-perceived status of cognition, fatigue and emotional functioning were reduced in patients (n ⫽ 30) compared to healthy controls (n ⫽ 10). Cognitive performance was impaired in patients with NHL compared to controls and a norm sample (n ⫽ 1179). PCCI was more severe in patients treated with rituximab and bendamustine (BR) than in patients who received R in combination with CHOP (cyclophosphamide, doxorubicin, vincristine, prednisone) polychemotherapy (R-CHOP). Individual alpha peak frequency and serum brain-derived neurotrophic factor (BDNF) levels in patients with NHL correlated with accuracy in the objective cognition test. Higher serum interleukin-6 (IL-6) concentrations were associated with higher fatigue levels. Patients with NHL and especially those who were treated with BR were affected by PCCI. BDNF and IL-6 might be involved in the pathogenesis of PCCI and fatigue. Keywords: Chemobrain, lymphoma, EEG, fatigue, cognition, cytokines
Introduction Post-chemotherapy cognitive impairment (PCCI), also known as “chemobrain,” is a well-known side effect of cancer therapies, affecting 17–75% of all patients with cancer [1]. Although PCCI focuses only on cognitive impairments
Correspondence: Philipp Zimmer, German Sport University Cologne, Department for Molecular and Cellular Sports Medicine, Am Sportpark Müngersdorf 6, 50933 Köln, Germany. Tel: 0049(0)221-4982-4821. Fax: 0049(0)221-4982-8370. E-mail: [email protected] Received 11 March 2014; revised 7 April 2014; accepted 12 April 2014
1
Leuk Lymphoma Downloaded from informahealthcare.com by Monash University on 12/04/14 For personal use only.
2
P. Zimmer et al.
et al. [7] and Kreukels et al. [8] combined EEG analyses with objective cognitive tests in patients with breast cancer to evaluate cognitive deficits. Their results indicated that the latency and amplitude of event related potentials are influenced by chemotherapy. These few studies have similar limitations to the studies discussed above. The infrequent application of EEG in the context of PCCI is remarkable, since Angelakis et al. [9] and Richard et al. [10] described a correlation between the individual alpha peak frequency (iAPF) and cognitive functions. The alpha rhythm is the dominant EEG pattern in the human scalp. iAPF is defined as the peak frequency within the range of the alpha frequency (7.5–12.5 Hz). For an extensive review see the article by Klimesch [11]. Rodin and Ahles [6] also made the criticism that objective and subjective cognition data are rarely combined with serum parameters when investigating PCCI. Although knowledge about the function of pro-inflammatory cytokines (i.e. interleukin-6 [IL-6] and tumor necrosis factor-α [TNF-α]) related to cognitive impairment is rudimental, they might be involved in generating and maintaining side effects such as fatigue and PCCI [12,13]. Besides pro-inflammatory cytokines, altered concentrations of neurotrophic factors such as brain-derived neurotophic factor (BDNF) may contribute to changes in cognitive function. BDNF is known to promote the growth and differentiation of neuronal stem cells and to support the survival of existing neurons. In addition, elevated BDNF levels are associated with enhanced long-term potentiation [14]. The aim of the present clinical trial was to determine the influence of chemoimmunotherapy in patients with non-Hodgkin lymphoma (NHL) on objective and subjective cognitive functions and potential (neuro-) physiological correlates. For this purpose, subjective and objective cognitive assessment tools were combined with EEG analyses and potential PCCI-relevant cytokine measurements in a group of patients with NHL and a healthy control group. In a subgroup analysis, patients who had been treated for indolent NHL with the monoclonal antibody rituximab and bendamustine (BR) were separated from those with aggressive NHL who had received R in combination with a poly-chemotherapy regimen, cyclophosphamide, doxorubicin, vincristine and prednisone (R-CHOP).
Participants and methods The study protocol was reviewed and approved by the local ethics committee in accordance with the Declaration of Helsinki. A trial registration was conducted at the German Clinical Trial Register, which is part of the world Health Organization (WHO) International Clinical Trials Registry Platform (No. DRKS0004743). All participants provide written informed consent prior to participation. Thirty-two patients were invited to participate. Two patients refused, leading to a recruitment rate of 93.75%.
Participants Over a period of 18 months, 30 patients with B-cell NHL (mean age 63.7 years, standard deviation [SD] ⫽ 12.8, 11
female, 19 male) and 10 healthy controls (mean age 56.6 years, SD ⫽ 10.7, three female, seven male) were recruited between March 2012 and September 2013. Of the 30 patients who received first-line therapy, 16 (five female, 11 male) were treated with BR whereas the other 14 patients (three female, 11 male) received R-CHOP. All patients were assessed within the first 3 months after their chemotherapy ended. The following inclusion criteria were used: first-line therapy with BR or R-CHOP had to be completed between 4 and 12 weeks prior to inclusion in the trial, and patients had to be older than 18 years. Patients who underwent second-line chemotherapy or radiation therapy or presented any condition (e.g. orthopedic restriction) preventing participation in the cognition tests were excluded.
Workflow Participants were examined once, between 9 a.m. and 10 a.m. After serum samples were taken, patients completed the European Organisation for Research and Treatment of Cancer Quality of Life Questionnaire-C30 (EORTC QLQ-C30), followed by a 5 min resting state EEG with closed eyes and a computer based cognition test.
Objective cognition assessment Executive function and attention were chosen as objective cognition parameters because these abilities have been described to be influenced by chemotherapy [4]. To determine the named abilities, we used the computer based Wiener test system (Schuhfried, Vienna, Austria). During the determination test (DT), patients had to react to visual and optical stimuli by pressing colored buttons on a keyboard and foot pedals for both feet. The test started after completing a standardized test instruction, including a practice period in order to ensure that the tasks were fully understood. All stimuli are presented adaptively. That is, the speed of presentation is adapted to the ability level of the participant. In detail, the DT can measure accuracy and processing speed under time pressure. According to the supervisory attentional system theory [15], the DT is an appropriate assessment to measure executive function. The main outcome is “correct responses.” This variable describes the individual processing limit under time pressure. Additional outcomes are “incorrect responses” and “omitted responses.” The internal consistency reliability for the variables ranges from r ⫽ 0.98 to r ⫽ 0.99. The results are presented as raw scores as well as percent rank compared to an age- and grade of education-matched norm sample, including a total number of 1179 healthy subjects. As described by the manufacturer, the majority of the norm sample has percent ranks between 25 and 75. Thus, percent ranks lower than 25 were considered as a weak performance, whereas percent ranks above 75 were considered as a strong performance [16].
Self-perceived status of cognition, fatigue and emotional functioning Subjective cognition status, fatigue and emotional functioning were evaluated using the subscales for cognitive function (CF), emotional function (EF) and fatigue (FA) of the EORTC
Chemobrain after B-cell non-Hodgkin lymphoma QLQ-C30 [17]. We decided to use this questionnaire because it includes all demanded domains. Additionally, this established questionnaire is widely used, making our data comparable to the outcomes of other studies.
Leuk Lymphoma Downloaded from informahealthcare.com by Monash University on 12/04/14 For personal use only.
EEG recording and data analysis Electroencephalography was recorded from 15 scalp locations using an electrode cap and a 16-channel EEG amplifier (Brain Products, Munich, Germany). Recording positions were equally distributed over the scalp (Fp1, Fp2, T7, F3, Fz, F4, T8, C3, Cz, C4, P3, Pz, P4, O1, O2) according to the international 10:20 system, with the electrical reference located on position FCz. Additionally, one electrooculographic electrode was used to detect vertical and horizontal eye movements and related artifacts in the EEG signal. Electrode impedances were kept below 5 kΩ. The sampling rate was set to 1000 Hz. For data analysis, Brain Vision Analyzer 2 software was used (Brain Products, Munich, Germany). An infinite-impulse response (IIR) band-pass filter (2–30 Hz, 48 dB/oct) was applied to the data to exclude low frequency artifacts as well as frequencies that were not relevant for the purpose of the study. In a second step, EEG data were re-referenced to the average potential of the 15 recording electrodes. Resting state data were then subdivided into segments of 4 s and baseline corrected. An independent component analysis (ICA) algorithm optimized for ocular artifact correction was applied, and those components representing ocular artifacts were excluded from further analysis. Subsequently, a semi-automatic artifact rejection (maximum allowed voltage step: 150 μV; maximal allowed voltage difference within 200 ms: 100 μV) was used to reject residual artifacts. Spectral power was calculated using fast Fourier transformation (FFT) (frequency resolution: 0.244 Hz) and frequency spectra were averaged across all segments. The iAPF was defined as the frequency indicating the highest power value within the 8–13 Hz range [18].
Serum parameter measurement IL-6 and BDNF concentrations of serum samples were analyzed using enzyme-linked immunosorbent assays
according to the manufacturer’s protocol (R&D systems, Minneapolis, MN).
Statistical analysis Statistical analysis was performed using SPSS 20 (IBM, Rochester, MN). In a first step, comparisons between the patient group and the control group were conducted with an unpaired t-test. Variance homogeneity was confirmed using the Levene test. In a second step, Spearman correlations were calculated for all variables. For subgroup analysis, analysis of variance (ANOVA) with subsequent Bonferroni post hoc test was performed. Furthermore, Spearman correlations were calculated for the subgroups of patients. To compare participants’ data for the objective cognition test with regard to those of the norm sample, median percent ranks were analyzed by Mann–Whitney U-test. The significance level was set at α ⫽ 0.05.
Results Participant characteristics In terms of age, no group differences could be detected (p ⫽ 0.123). According to the disease incidence rate, sex distribution was not equal among the groups (see “Participants and methods” above).
Objective cognition assessment outcomes The t-test revealed significant group differences for the main outcome of the objective cognition test (“correct responses”). The number of “correct responses” was significantly lower in patients when compared to the control group (p ⫽ 0.003). “Incorrect responses” and “omitted responses” did not differ significantly between the groups (Table I). Patients who were treated with BR achieved significantly lower numbers of “correct responses” than the control group (p ⬍ 0.001) and the R-CHOP group (p ⫽ 0.005). Additionally, patients who received BR had significantly more “omitted responses” than the control group (p ⫽ 0.034). The outcomes of the R-CHOP group did not differ from those of the healthy controls (Table II).
Table I. Results of all outcomes of patient and control groups†. Control group Objective cognition test outcomes Correct responses 228 (7.35) Incorrect responses 13.5 (4.10) Omitted responses 12.2 (2.42) Subjective perception parameters Cognition 88.3 (4.34) Fatigue 9.96 (4.20) Emotional function 81.5 (6.03) EEG data iAPF frontal region 9.26 (0.30) iAPF central region 9.41 (0.28) iAPF parietal region 9.47 (0.25) Cytokines IL-6 (pg/mL) 1,94 (0.19) 15.40 (3.01) BDNF (μg/mL)
3
Patients
p-Value
95% CI
186 (7.05) 16.1 (2.11) 18.4 (1.76)
0.003* 0.412 0.060
[⫺ 67.8; ⫺ 15.3] [⫺ 6.1; 11.4] [⫺ 0.27; 12.6]
64.4 (5.04) 54.4 (4.70) 58.5 (5.26)
0.013* 0.000* 0.024*
[⫺ 42.4; ⫺ 5.40] [31.64; 57.33] [⫺ 42.9; ⫺ 3.16]
9.66 (0.09) 9.44 (0.09) 9.99 (0.15)
0.193 0.844 0.093
[0.23; 1.01] [⫺ 0.61; 0.68] [⫺ 0.88; 1.13]
3,67 (0.33) 14.23 (1.20)
0.000* 0.723
[0.95; 2.49] [⫺ 81.8; 58.4]
EEG, electroencephalographic; iAPF, individual alpha peak frequency; IL-6, interleukin-6; BDNF, brain-derived neurotrophic factor. *Significant difference, t-test. †Mean, standard error of the mean and 95% confidence interval (CI) are presented for all outcomes.
4
P. Zimmer et al. Table II. Results of all outcomes of subgroup analysis. ANOVA
Leuk Lymphoma Downloaded from informahealthcare.com by Monash University on 12/04/14 For personal use only.
Objective cognition test outcomes Correct responses ⬎ 0.001* Incorrect responses 0.837 Omitted responses 0.027* Subjective perception parameters Cognition 0.006* Fatigue ⬍ 0.001* Emotional function 0.078 EEG data iAPF frontal region 0.040* iAPF central region 0.704 iAPF parietal region 0.185 Cytokines IL-6 (pg/mL) 0.020* 0.258 BDNF (μg/mL)
Controls vs. R-CHOP 0.354 1.00 1.00
Controls vs. BR 0.000* 1.00 0.034*
R-CHOP vs. BR 0.005* 1.00 0.167
0.005* ⬍ 0.001* 0.108
0.359 ⬍ 0.001* 0.165
0.137 1.00 1.00
1.00 1.00 0.757
0.051 1.00 0.204
0.219 1.00 1.00
0.051 1.00
0.028* 0.852
1.00 1.00
EEG, electroencephalographic; iAPF, individual alpha peak frequency; IL-6, interleukin-6; BDNF, brain-derived neurotrophic factor; ANOVA, analysis of variance; R-CHOP, rituximab, cyclophosphamide, doxorubicin, vincristine, prednisone; BR, rituximab, bendamustine. *Significant difference, p-value, ANOVA and Bonferroni post hoc.
The median percent rank of the outcomes “correct responses” and “omitted responses” showed that patients who were treated with BR had a median percent rank lower than 25. Patients of the R-CHOP group as well as the control group revealed a median percent rank within the usual range (25–75) of the norm sample (Figure 1). Subjects of the control group showed a significantly higher median percent rank regarding “correct responses” than the patient group and both patient subgroups. No group differences could be detected for “incorrect responses.” Median percent rank for “omitted responses” differed significantly between controls and patients as well as the BR subgroup (Figure 1).
Self-perceived outcomes Patients demonstrated significantly lower scores of subjective perception of cognition, fatigue and emotional functioning (Table I). Within the subgroup analysis, patients in the R-CHOP group reported significantly more cognitive impairments than the control group (p ⫽ 0.005), whereas no difference was found between the other groups. Fatigue data did not differ between the patient groups but were significantly elevated in comparison to the control group (Table II).
EEG data iAPF data did not differ between patients and controls. ANOVA of the frontal iAPF yielded a significant group effect (p ⫽ 0.040). The subsequent post hoc test showed a trend toward lower frontal iAPF values in the BR group compared to controls. However, this difference did not reach statistical significance (p ⫽ 0.051) (Table II).
Cytokine measurements IL-6 serum levels were significantly elevated in the patient group (p ⬍ 0.001) (Table I). The BR group exhibited significantly higher IL-6 serum concentrations than the healthy control group. The R-CHOP group also showed elevated IL-6 serum concentrations. However, the difference from the control group did not reach significance (p ⫽ 0.051) (Table II). BDNF levels did not show any group differences.
Correlations All significant correlations are listed in Table III. IL-6 serum concentrations correlated with fatigue levels (p ⫽ 0.003, r ⫽ 0.522) and age (p ⫽ 0.026, r ⫽ 0.406), but did not show any correlation with outcomes of the objective cognition test. Patients with higher BDNF levels had better results in the main variable (“correct responses”) of the DT (p ⫽ 0.041,
Figure 1. Boxplots of median percent rank for main variables of the objective cognition test for all groups and subgroups. Significant differences of the median percent ranks are presented with p-values. Percent ranks ranging from 25 to 75 are regarded as normal (gray shading represents data of the norm sample). Percent rank smaller than 25 is seen as a weak development of the variable. In addition to medians, means are indicated by ⫹.
Chemobrain after B-cell non-Hodgkin lymphoma
5
Leuk Lymphoma Downloaded from informahealthcare.com by Monash University on 12/04/14 For personal use only.
Table III. Significant correlations of all outcomes of patient group and both subgroups*. Patient group (n ⫽ 30) IL-6 BDNF iAPF parietal region iAPF frontal region Age Age Age BR (n ⫽ 16) IL-6 iAPF central region iAPF central region iAPF parietal region iAPF frontal region iAPF frontal region Age Age R-CHOP (n ⫽ 14) iAPF frontal region iAPF parietal region
Outcome
p-Value, correlation coefficient
Fatigue Correct responses Omitted responses Emotional functioning Correct responses Subjective perception of cognition IL-6
p ⫽ 0.003, r ⫽ 0.522 p ⫽ 0.041, r ⫽ 0.375 p ⫽ 0.017, r ⫽ ⫺ 0.432 p ⫽ 0.027, r ⫽ ⫺ 0.403 p ⫽ 0.008, r ⫽ ⫺ 0.474 p ⫽ 0.001, r ⫽ 0.586 p ⫽ 0.029, r ⫽ 0.428
Fatigue Correct responses Incorrect responses Omitted responses Emotional functioning Subjective perception of cognition Subjective perception of cognition IL-6
p ⫽ 0.034, r ⫽ 0.532 p ⫽ 0.033, r ⫽ 0.534 p ⫽ 0.025, r ⫽ ⫺ 0.558 p ⬍ 0.001, r ⫽ ⫺ 0.773 p ⫽ 0.024, r ⫽ ⫺ 0.559 p ⫽ 0.047, r ⫽ ⫺ 0.503 p ⫽ 0.018, r ⫽ 0.581 p ⫽ 0.039, r ⫽ 0.601
Incorrect responses BDNF
p ⫽ 0.023, r ⫽ 0.601 p ⫽ 0.010, r ⫽ ⫺ 0.633
EEG, electroencephalographic; IL-6, interleukin-6; BDNF, brain-derived neurotrophic factor; iAPF, individual alpha peak frequency; BR, rituximab, bendamustine; R-CHOP, rituximab, cyclophosphamide, doxorubicin, vincristine, prednisone. *All significant correlations between objective cognition test, subjective perception, EEG data and serum parameters of patient group and subgroups.
r ⫽ 0.375). We found correlations between the iAPF of the parietal region and “omitted responses” (p ⫽ 0.017, r ⫽ ⫺ 0.432) and the iAPF of the frontal cortex and emotional functioning (p ⫽ 0.027, r ⫽ 0.403) in the DT. A similar pattern of correlation was obtained in the BR group (with the exception of the BDNF data). Additionally, in the BR group, significant correlations were found between all outcomes of the objective cognition test and the iAPF data. In contrast to the BR group, the described correlations were not detectable in patients who received R-CHOP.
Discussion This study is the first comprehensive analysis of the impact of chemoimmunotherapy on cognitive functions of patients with B-cell NHL combined with serum and neurophysiological parameters. Furthermore, this is the first study comparing two different chemotherapy strategies (BR vs. R-CHOP) regarding their impact on cognitive abilities and associated physiological outcomes in patients with NHL. Our results indicate that patients with NHL have cognitive impairments within 3 months after the end of their chemoimmunotherapy. These results are in accordance with the patients’ subjective perception of cognition and with comparable studies [6,19]. Subgroup analyses revealed that patients treated with BR showed significantly lower performance scores in the objective cognition test than controls, the R-CHOP group and the age-matched norm sample. Indeed, one case report indicates that the application of BR is associated with a cognitive decline [20]. However, we expected that patients treated with R-CHOP, a regimen including one neurotoxic agent (vincristine) and one blood–brain barriercrossing drug (prednisone), would have elevated cognitive impairments. In contrast to Tholstrup et al. [21], we found an impact on the subjective perception of cognition after chemoimmunotherapy with R-CHOP. These findings were
not confirmed by the outcomes of the objective cognition assessment. Differences between controls and patients with NHL regarding the iAPF were not detected. However, a trend toward lower iAPF levels in the BR group was observed. Significant correlations of the iAPF and objective cognition outcomes were found within patients, and especially within the BR subgroup. These results confirmed other studies reporting a correlation of high iAPF values and selective aspects of cognition [9,10]. The detected correlation of frontal iAPF values and emotional functioning is in accordance with findings in depressive patients [22]. Therefore, regional iAPF measurements might be used for the objective evaluation of both cognitive impairments and emotional disorders. BDNF serum levels correlated significantly with the main outcome (“correct responses”) of the cognition test. These results confirm those of Oral et al. [23], who found reduced BDNF concentrations in depressive and cognitively impaired patients. We assume that serum BDNF levels may be used as an additional surrogate parameter to supplement the results of objective cognitive assessments. Additionally, they could help to understand the underlying mechanism of PCCI. The correlation of IL-6 serum levels and fatigue is in agreement with other studies, leading to the hypothesis that inflammation is a main cause of fatigue, independent of cancer type and therapy [12,13,24]. Elevated inflammation markers were also detected in various patient groups with depressive disorders, which would support the connection with fatigue and depression [25]. Further evidence for this relationship between inflammation and fatigue comes from animal studies [26]. However, the translational character of such findings is restricted. Future research in this field should focus on inflammation markers such as IL-6 within the brain and their origin. The use of IL-6 serum level as a single marker for these symptoms might not be sufficient. Inconspicuous IL-6 concentrations in patients in the R-CHOP group
Leuk Lymphoma Downloaded from informahealthcare.com by Monash University on 12/04/14 For personal use only.
6
P. Zimmer et al.
were not surprising, since the therapy protocol included the anti-inflammatory drug prednisone. The results of the present study should be considered within the context of its limitations. First, patients were not tested in a longitudinal design. Since numerous studies have shown cognitive impairments prior to chemotherapy, the phenomenon of PCCI might be biased by typical pretreatment symptoms such as depression, anxiety and stress [27,28]. Regarding the self-perceived status of cognition, fatigue and emotional functioning, we decided to use the well-established and often used EORTC QLQ-C30. The named domains of this questionnaire constitute only a few questions. One could use more detailed questionnaires; however, it should be taken into account that study subjects, and especially those with cognitive impairments, may have problems with completing a vast number of questionnaires. As objective assessments, other cognitive abilities such as memory should be included in future studies. Finally, studies could use a wider range of biomarkers in larger patient groups. In conclusion, our results reveal that patients with NHL, in particular when treated with BR, show cognitive impairments within 3 months after their chemoimmunotherapy. Objective cognitive assessments are necessary to detect PCCI. iAPF, as a neuroelectric index of cognition, as well as BDNF and IL-6 serum levels are biomarkers that deserve further study in the context of PCCI and fatigue. EEG analysis as well as BDNF levels might have the potential to support results from cognition tests and enable us to learn more about the underlying mechanisms.
Acknowledgements The authors would like to thank Dr. Pauline Schuhmacher and Mrs. Muriel Freudenberger for editorial support. Dr. Schuhmacher is supported by the German Federal Ministry of Research and Education (BMBF grant 01KN1106). Mrs. Freudenberger is supported by the Cologne Excellence Cluster for Cellular Stress Responses in AgingAssociated Diseases (CECAD) (DFG EXC 229). Furthermore, the authors would like to thank Mrs. Anke Schmitz and Mrs. Anika Voss for their technical support. Potential conflict of interest: Disclosure forms provided by the authors are available with the full text of this article at www.informahealthcare.com/lal.
References [1] Myers JS. Chemotherapy-related cognitive impairment: neuroimaging, neuropsychological testing, and the neuropsychologist. Clin J Oncol Nurs 2009;13:413–21. [2] van Dam FS, Schagen SB, Muller MJ, et al. Impairment of cognitive function in women receiving adjuvant treatment for high-risk breast cancer: high-dose versus standard-dose chemotherapy. J Natl Cancer Inst 1998;90:210–218. [3] Scherling CS, Smith A . Opening up the window into “chemobrain”: a neuroimaging review. Sensors (Basel) 2013;13:3169–3203. [4] Hodgson KD, Hutchinson AD, Wilson CJ, et al. A meta-analysis of the effects of chemotherapy on cognition in patients with cancer. Cancer Treat Rev 2013;39:297–304.
[5] Hutchinson AD, Hosking JR, Kichenadasse G, et al. Objective and subjective cognitive impairment following chemotherapy for cancer: a systematic review. Cancer Treat Rev 2012;38:926–934. [6] Rodin G, Ahles TA . Accumulating evidence for the effect of chemotherapy on cognition. J Clin Oncol 2012;30:3568–3569. [7] Schagen SB, Hamburger HL, Muller MJ, et al. Neurophysiological evaluation of late effects of adjuvant high-dose chemotherapy on cognitive function. J Neurooncol 2001;51:159–165. [8] Kreukels BP, Schagen SB, Ridderinkhof KR, et al. Electrophysiological correlates of information processing in breast-cancer patients treated with adjuvant chemotherapy. Breast Cancer Res Treat 2005;94:53–61. [9] Angelakis E, Lubar JF, Stathopoulou S. Electroencephalographic peak alpha frequency correlates of cognitive traits. Neurosci Lett 2004;16;371:60–63. [10] Richard Clark C, Veltmeyer MD, Hamilton RJ, et al. Spontaneous alpha peak frequency predicts working memory performance across the age span. Int J Psychophysiol 2004;53:1–9. [11] Klimesch W. EEG alpha and theta oscillations reflect cognitive and memory performance: a review and analysis. Brain Res Brain Res Rev 1999;29:169–195. [12] Seruga B, Zhang H, Bernstein LJ, et al. Cytokines and their relationship to the symptoms and outcome of cancer. Nat Rev Cancer 2008;8:887–899. [13] Ganz PA , Bower JE. Cancer related fatigue: a focus on breast cancer and Hodgkin’s disease survivors. Acta Oncol 2007;46: 474–479. [14] Huang EJ, Reichardt LF. Neurotrophins: roles in neuronal development and function. Annu Rev Neurosci 2001;24:677–736. [15] Norman DA , Shallice T. Attention to action: willed and automatic control of behaviour. In: Shapiro DL, Schwartz G, editors. Consciousness and self-regulation: advances in research. New York: Plenum Press ; 1986. pp 1–14. [16] Neuwirth W, Benesch M. Manual determination test. 35 ed. Mödling, Austria: Schuhfried GmbH; 2012. [17] Aaronson NK , Ahmedzai S, Bergman B, et al. The European Organization for Research and Treatment of Cancer QLQ-C30: a quality-of-life instrument for use in international clinical trials in oncology. J Natl Cancer Inst 1993;85:365–376. [18] Doppelmayr M, Kliemesch W, Pachinger T, et al. Individual differences in brain dynamics: Important implications for the calculation of event related band power. Biol Cybern 1998;79:49–57. [19] Jensen RE, Arora NK, Bellizzi KM, et al. Health-related quality of life among survivors of aggressive non-Hodgkin lymphoma. Cancer 2013;119:672–680. [20] Küllmer A, Herrmann G, Riecken B. [Rapid-progressive decrease of cognitive and physical functions in a B-cell non-Hodgkin’s lymphoma patient treated with rituximab/bendamustine]. Dtsch Med Wochenschr 2013;138:1355–1359. [21] Tholstrup D, Brown Pde N, Jurlander J, et al. Quality of life in patients with diffuse large B-cell lymphoma treated with dosedense chemotherapy is only affected temporarily. Leuk Lymphoma 2011;52:400–408. [22] Arns M, Drinkenburg WH, Fitzgerald PB, et al. Neurophysiological predictors of non-response to rTMS in depression. Brain Stimul 2012;5:569–576. [23] Oral E, Canpolat S, Yildirim S, et al. Cognitive functions and serum levels of brain-derived neurotrophic factor in patients with major depressive disorder. Brain Res Bull 2012;88:454–459. [24] Bower JE, Ganz PA , Irwin MR, et al. Inflammation and behavioral symptoms after breast cancer treatment: do fatigue, depression, and sleep disturbance share a common underlying mechanism? J Clin Oncol 2011;29:3517–3522. [25] Elderkin-Thompson V, Mintz J, Haroon E, et al. Executive dysfunction and memory in older patients with major and minor depression. Arch Clin Neuropsychol 2007;22:261–270. [26] Patki G, Solanki N, Atrooz F, et al. Depression, anxiety-like behavior and memory impairment are associated with increased oxidative stress and inflammation in a rat model of social stress. Brain Res 2013;1529:73–86. [27] Wefel JS, Lenzi R, Theriault RL, et al. The cognitive sequelae of standard-dose adjuvant chemotherapy in woman with breast carcinoma: results of a prospective, randomized, longitudinal trial. Cancer 2004;100:2292–2299. [28] Jansen CE, Cooper BA , Dodd MJ, et al. A prospective longitudinal study of chemotherapy-induced cognitive changes in breast cancer patients. Support Care Cancer 2011;10:1647–1656.
