Psychosocial Study

Cognitive function in the acute course of allogeneic hematopoietic stem cell transplantation for hematological malignancies


The aim of the study was to assess cognitive performance in patients with hematological malignancies before, and 3 months after, allogeneic hematopoietic stem cell transplant (HSCT). A consecutive sample of 39 patients was assessed before admission with a comprehensive neuropsychological test battery and health-related quality-of-life (HRQoL) questionnaires; 19 of these patients were retested around 100 days post HSCT. Test results were compared with normative data and revealed minimal differences at both time points in the level of group-means. One parameter – simple reaction time – was significantly worse (prolonged) at second measurement after HSCT. According to the definition of an impairment score (more than three impaired functions), 26% of patients were classified as impaired before as well as after HSCT. Neuropsychological test results did not vary systematically according to medical variables such as extent of pretreatment, graft-versus-host-disease (GvHD) and kind of conditioning protocol. As a dimension of HRQoL, self-rated cognitive function was in the normal range before and after HSCT. Significant correlations between HRQoL and neuropsychological parameters were related to symptom scales. This study showed impairments of neuropsychological performance for a subgroup of patients before and after allogeneic HSCT. Systematic effects of conditioning, medical variables or self-rated HRQoL could not be observed.


Since chemotherapy combined with hematopoietic stem cell transplantation (HSCT) has become accepted standard therapy for hematological malignancies and overall survival has risen, there is growing interest in cognitive functioning as well as in long-term quality of life (QoL) of transplanted patients.1, 2, 3, 4, 5

The vast majority of studies in this field have been conducted in cross-sectional designs. On this account, questions about the course of, and whether there are limitations in cognitive functioning even before the beginning of, any chemotherapy have been widely discussed.6, 7, 8 Wefel et al.7 found that 35% of 86 patients with breast carcinoma ‘demonstrated impaired overall cognitive function before the initiation of adjuvant chemotherapy’ (p 471). These authors criticize the fact that many cross-sectional studies failed to report on performance before therapy and therefore overestimated the extent of cognitive impairment.

Data concerning allogeneic HSCT and cognitive performance are sparse. Harder et al.9 studied a population of hematology patients 22–82 months (mean: 45 months) after transplant – 87.5% of these were allogeneic. Functions most frequently impaired were selective attention and executive functioning, information processing speed, visual memory and verbal learning. In all, 60% of participants performed at least one test in the impaired range. Predictors for impairment were QoL (fatigue, global health) rather than medical data (i.e., complications of the central nervous system (CNS), graft-versus-host-disease (GvHD) and its treatment).

Padovan et al.10 assessed HSCT patients at a mean of 34 months after treatment. A total of 37% of allogeneic HSCT patients showed a conspicuous neuropsychological profile, mainly concerning memory functions. In this study, risk factors for impairment were cyclosporine, age over 40 years and chronic GvHD.

These studies were conducted retrospectively and no causal relationships could be identified, but two recent papers have followed allogeneic patients samples prospectively.

Sostak et al.11 studied 71 patients – most of whom had received total body irradiation (TBI) – 2–4 months before and 14 (+/−3) months after allogeneic HSCT. The authors report a high percentage of neurologic (abnormalities in 65% of 71 participating patients) and neuroradiologic (pathologic magnetic resonance imaging in 56% of 55 patients examined) abnormalities before as well as after SCT. Concerning neuropsychological screening, 58% of participants had at least one abnormal test result before SCT, reducing to 51% after treatment.

In a prospective study published by Syrjala et al.12 142 myeloablative allogeneic transplants recipients were included. Fifty-four patients were tested at three points in time: before admission, and 80 days and 1 year following HSCT. Using standardized neuropsychological instruments covering memory, attention, executive and psychomotor functions, the authors reported a consistent trend of decline of functions in the first 80 days after transplant and a further trend of returning back to pretransplant functioning in all areas except grip strength and motor dexterity 1 year after treatment. Nonetheless, there were significant numbers of impaired patients at all time points, again with a trend of decline at day 80 and amelioration at 1 year (45% pre-transplant, 78% 80 days post-transplant and 53% 1 year post-transplant with two and more impairments).

Looking at the literature concerning cognitive deficits in the course of chemotherapy and HSCT, there are only limited data shedding light on the questioned whether these treatments affect cognitive performance at all.3, 13, 14 Contradictory results exist concerning the extent of impairments either before, or after chemotherapy. Finding deviations from norms before treatment leads to the discussion of other factors not directly associated with the treatment such as influence of the malignancies themselves or psychological factors (anxiety, treatment-related distress, depression). Finally, aspects of subjectively rated QoL, such as physical functioning and fatigue, or objective treatment associated side effects could add to the observation that patients may not reach a cognitive performance in the normal range.

Anderson-Hanley et al.13 conducted a meta-analysis of the neuropsychological effects of cancer treatment assuming that the ‘data are consistent with the idea that there is an adverse effect’ of chemotherapies. Nonetheless, they point to the fact that these effects are much stronger in studies using cross-sectional between-subject designs, compared to those with longitudinal designs. Future research has to ‘incorporate such designs or the data will add little to what we already know’ (p 978).

The following pilot study therefore aimed to identify prevalences of cognitive impairment in the course of HD/HSCT in a homogenous sample of patients with hematological malignancies, and to analyze associations between cognitive impairment and medical characteristics and dimensions of HRQoL. Thereby, this study adds longitudinal data of patients receiving myeloablative and non-myeloablative conditioning.

Patients and methods

Study design and participants

This pilot study is based on a prospective design approved by the ethics committee of the Hamburg physicians' chamber. Patients with newly diagnosed hematological malignancies scheduled for allogeneic transplant of hematopoietic stem cells at the transplant unit of the Hamburg University Medical Center were consecutively recruited once at a maximum of 2 weeks before admission (baseline, T0) and 100 days (range: 80–120 days) after transplant (T1) in the outpatient department. Conditioning started the following day, or 2 days after admission, and lasted between 5 and 8 days. Day 100 was chosen, because patients were expected to be free of the most severe acute treatment side effects at this point in time. All patients received either standard or reduced intensive conditioning.

Written informed consent was obtained from all participants. The inclusion criteria were an age of at least 18 years and the ability to speak and read the German language fluently. Exclusion criteria – comprising factors that could influence cognitive functioning independently from conditioning – were the following: presence of a solid tumor, second HSCT, intrathecal therapy, cranial radiation, intake of psychotropic drugs, barbiturates or morphine, central nervous system disease (CNS) disease or neurological symptoms potentially influencing cognitive performance, history of – or current – psychiatric disease including substance abuse and uncorrected vision.

Study group

During the study period, 87 patients were eligible; 39 (45%) of these completed neuropsychological testing and completed questionnaires. Twenty-seven patients (31%) did not participate, mainly for organizational and health-related reasons (Table 1) and a further 21 patients (24%) completed questionnaires but could not be tested.

Table 1 Response rates at baseline and T1, and reasons for non-participation

Between T0 and T1 (100 days after HSCT), nine patients died, so that 30 patients were candidates for the second assessment. Ten (33.3%) of these did not participate for organizational or health-related reasons, and one patient declined. The remaining 19 patients (63.3% of eligible patients at T1) were assessed with both neuropsychological tests and self-report questionnaires (Table 1).

Measures and procedure

The pre-transplant and follow-up evaluation included a health-related quality-of-life (HRQoL) questionnaire and a neuropsychological examination. The total assessment lasted 2 h at most and was conducted by trained and experienced neuropsychologists. In addition, patients filled out questionnaires concerning demographic and medical data. Diagnostic and treatment information were obtained by medical chart review.

The neuropsychological test battery consisted of standardized and normed paper and pencil tests as well as computer-based measures. Tests were selected for reliability, validity and sensitivity to cover the global domains of memory, attention and executive functions. Each domain contained several specific cognitive functions assessed by one or two neuropsychological tests or subtests. This procedure led to the following structure: three cognitive domains divided into 12 specific functions, represented by 19 test parameters (Table 2; see legend for detailed description of tests).

Table 2 Neuropsychological assessment battery

Health-related quality of life was assessed using the European Organization for Research and Treatment of Cancer (EORTC). Quality of Life Core Questionnaire (QLQ-C3024), a self-report questionnaire designed for adult cancer patients. The QLQ-C30 includes five functional scales, three symptom scales and a global health status scale as well as six single items. The quantitative scores range from 0 to 100; regarding functional scales, higher scores represent better functioning, whereas higher scores in symptom scales and items reflect a greater number of complaints.

Statistical analysis and definition of neuropsychological impairment

The Statistical Package for the Social Sciences (SPSS) for Windows, version 12.0, was used for all data analyses. To facilitate interpretation, neuropsychological raw-scores were z-transformed according to age-corrected test norms. To rate test performance as impaired/not impaired, the following cutoff scores according to Spreen and Strauss25 were used: z>−1.4=‘not impaired’; z−1.4=‘impaired’. Furthermore, for each patient, an impairment score (IS; ranging from 0 to 12) as a sum of impaired functions was calculated. Concerning functions being represented by more than one test, one impaired test was sufficient for impairment of function. An IS of more than 3 led to a global rating of ‘impaired’.

Nonparametric or parametric tests were used according to normal distribution of variables and scale level – the assumption of normal distribution was proven with Kolmogorov–Smirnov tests. To compare patients who were assessable neuropsychologically at both points in time with the remainder at T0, t-tests were applied for age, χ2-tests were applied for gender, Mann–Whitney U-tests for extent of pretreatment and education, t-tests and Mann–Whitney U-tests, for QLQ-C30 scores. Differences between T0 and T1 concerning quality-of-life data were tested according to normal distribution of scales, with t-tests and Wilcoxon tests for dependent samples. Significant deviations from population norms in neuropsychological tests were detected by setting the mean 0 and conducting single t-tests for every parameter. Differences in z-transformed means of neuropsychological data were analyzed using t-tests for dependent samples, whereas frequencies of impairments at T0 and T1 were compared with McNemar tests. Relations between neuropsychological test results and quality-of-life (QoL) variables were assessed by calculating Pearson's correlation coefficients. To adjust for multiple comparisons, a 0.01 level of significance was used concerning pre–post comparisons, test-differences from norms and correlations with QoL variables


Sample characteristics

Table 3 summarizes demographical and medical data for subjects only interviewed at T0 (N=21), for those interviewed and tested at T0 (N=39) and for those interviewed and tested at T0 and T1 (N=19). The 39 patients who were fully assessed before admission were on average 45.6 years old (range 22–69), 64.1% were male. Education distribution was comparable in this group as about one-third had not attended high school, one-third had attended high school and one-third had had college education.

Table 3 Demographical and medical data at T0 and T1

Distribution of diagnoses reflected a wide spectrum of hematological diseases: nearly half of the fully assessed sample (46.2%) was transplanted for acute myeloic leukemia or chronic myeloic leukemia, multiple myeloma, lymphoma, myelodysplastic syndrome and osteomyelofibrosis. Each constituted about 10% of patients. On admission for HSCT, extent of chemotherapy pretreatment was rated by the admitting hematologist (rating: no/little: no hemato-oncological pretreatment at all or at most, long-term oral treatment; medium: at most one cycle of chemotherapy; extended/strong: several cycles of chemotherapy or high-dose therapy with autologous HSCT): 15 patients (38.5%) received no or little, 13 (33.3%) medium and 11 (28.2%) extended therapies before HSCT.

Table 4 displays features of medical treatment: of the 19 participants at T1, seven (36.8%) received related, 12 (63.2%) unrelated transplants. The same numbers were observed for incidence of at least grade II acute GvHD (grading according to Glucksberg et al.26): 12 developed GvHD, seven did not (see Table 4).

Table 4 Medical data at T1 (N=19)

Conditioning was conducted following standard dose treatment (ST) in seven cases, and 12 patients received reduced intensive conditioning (RIC). There were no significant correlations between conditioning or kind of transplant and development of GvHD (seven RIC patients vs five ST patients and nine unrelated vs three related transplants developed GvHD). Conditioning protocols contained six different chemotherapy agents (see Table 4), with busulfan or treosulfan as major components. Other substances were cyclophosphamide or etoposide, mainly for leukemias or lymphomas, and melphalan or fludarabine, mainly for multiple myeloma. The majority of transplants (N=17) received anti-thymocyte globulin (ATG) before allogeneic transplant. Two conditioning regimens included TBI at 2 (RIC) and 12 (ST) Gy, respectively.

Exclusion of systemic differences

Because of the high attrition rate and to exclude systemic differences, we conducted several analyses comparing patients only interviewed (N=21) at T0 with those fully assessed (N=39) and the latter group with the 19 patients who had participated in both neuropsychological assessments (N=19). Comparisons of the two sub-samples at T0 revealed no significant differences with respect to age (P=0.50), gender (P=0.57), education (P=0.46), diagnosis (P=0.32) and pretreatment (P=0.12). This was equally true for patients interviewed and tested only at T0 and those fully assessed at both points in time (age: P=0.65; gender: P=0.58; education: P=0.61; diagnoses: P=0.91; pretreatment: P=0.68). The same analysis was carried out for the EORTC-QLQ-C30 scales and z-transformed neuropsychological parameters. Again, no significant differences between the aforementioned groups could be detected.

Health-related quality of life

Details of quality-of-life data for participants at T0 and T1 are listed in Table 5. Only one scale differed on a 0.01 level of significance between T0 and T1: physical functioning declined from a mean of 87–80 (P=0.001). Self-rated cognitive functioning remained high and almost stable at both points in time (M=85 at T0 and M=86 at T1). Concerning symptom scales, only nausea/vomiting increased significantly from a mean of 2–13 (P=0.01).

Table 5 Health related quality of life at T0 and T1

Neuropsychological test results

Deviations from population norms and differences between T0 and T1

Table 6 shows z-transformed test-mean scores, standard deviations and P-values of comparisons against population norms for samples at T0 and T1. At T0, two of the observed means differed significantly at P0.01 from the population parameter: Verbal long-term memory (measured with Verbaler Lern- und Merkfähigkeitstest (VLMT) trials 5 minus 6) was below the norm population (P=0.010), whereas reasoning (according to the Leistungs-Prüf-System (LPS)4) was significantly better (P<0.001) than the norm.

Table 6 Norm comparisons and differences in z-transformed mean scores across time

At 100 days post HSCT, another test parameter of verbal long-term memory (VLMT trials 5 minus 7) was below norm (P=0.001). Additionally, visual working memory measured with the visual memory span subtest of the Wechsler Memory Scale showed a mean below norm (P=0.002); again reasoning exceeded population norms significantly (P<0.01 for LPS3 and P<0.001 for LPS4).

Analyzing potential differences between assessments, most parameters did not differ significantly at the 0.01 P-level between assessments. A significant decline in cognitive performance was observed in simple reaction time (P=0.01).

Impairments concerning single functions and global impairment

Results for impaired and non-impaired tests are shown in Table 7. Following our criterion of impairment (1.4 SD), between one and five patients (5.3–26.3%) were impaired before admission. In all, 1–8 patients corresponding with 5.3–42.1% showed impaired test results at day 100 after HSCT. Regarding executive functions, no patient displayed impairment of reasoning at either point in time. In terms of differences between proportions of impaired and non-impaired individuals, none of the observed frequencies of impairment between T0 and T1 were statistically significant.

Table 7 Frequencies of impairments at T0 and T1 and P-values of McNemar tests across time

The IS as the sum of abnormal cognitive functions could range from 0 to 12; Figure 1 displays impairments for T0 and T1. In all, 21.1% of patients demonstrated no cognitive impairment at T0 and T1. Ten patients (52.6%) showed impairments in 1–3 functions (‘mild impairment’) at both assessments. According to our definition of ‘global impairment’ – four and more abnormal functions – five patients (26.3%) were rated as globally cognitively impaired at T0; this percentage was found to be exactly the same at T1. Regarding intraclass changes, 17 of 19 patients showed stable performance and were ranging in the same class of impairment (no, mild, global) at both time points, one patient improved from ‘global’ to ‘mild’ impairment, another one declined from ‘mild’ to ‘global’ impairment.

Figure 1

Percentages of impairments at T0 and T1 – Legend: Cognitive performance is rated as ‘globally impaired’ if more than three functions are rated as impaired

Medical features, quality of life and neuropsychological performance

There was no single significant difference in z-mean scores of neuropsychological tests between patients experiencing at least grade II GvHD and those with no or grade I GvHD. Regarding extent of pretreatment (no, medium, extended), again no single difference between groups reached statistical significance. Comparing results of related vs unrelated transplants, differences were significant concerning the digit-span backwards task (P=0.010). Inspection of group means revealed better results for recipients of related transplants.

Because this study is, to our knowledge, the first to compare patients undergoing myeloablative and non-myeloablative conditioning, we display test results of subjects receiving standard and reduced-intensive conditioning in Table 8. One single significant result shows a better performance of RIC patients on the reasoning task LPS-4 (P=0.01). As a trend, the opposite is true for selective attention (d2), with RIC patients having worse results than subjects undergoing standard conditioning (P=0.03).

Table 8 Z-transformed test mean scores and standard deviations for patients undergoing standard or reduced intensive conditioning

Looking for relationships between features of health-related QoL and neuropsychological performance at both assessments, no significant correlations between functional QoL scales and cognitive performance were observed. Regarding symptom scales at T0, three coefficients reached significance on a 0.01 P-level: Verbal working memory (VLMT) correlated negatively with short breath (r=−0.598; P=0.007), phasic alertness (Testbattery for Attentional Performance) with nausea (r=−0.636; P=0.003) and word fluency (Regensburg Word Fluency Test) with financial difficulties (r=−0.594; P=0.007). At T1, information processing speed (Trailmaking test A) had a negative correlation with nausea (r=−0.703; P=0.001) and again word fluency with financial difficulties (r=−0.644; P=0.003).


This study aimed to analyze cognitive performance during the acute course of hematological therapy and hematopoietic stem cell transplantation, including both myeloablative and non-myeloablative conditioning. Patients with hematological malignancies were assessed twice with a comprehensive neuropsychological test battery and a questionnaire concerning HRQoL. Thirty-nine patients completed data sets before admission and 19 patients were able to complete both assessments.

Looking at the results of neuropsychological tests, we saw the following pattern: one parameter at T0 and two parameters at T1 were below population norms – both were related to the memory domain. Executive functions such as reasoning were even better than in the normal population. One basic function from the attention domain – the simple reaction time – was worse after HSCT compared with the first assessment.

Besides comparing group means, we rated neuropsychological performance in every single test as impaired or non-impaired. We chose a comparatively conservative criterion: each result had to be at least 1.4 standard deviations below the mean of the norm population to be rated as ‘impaired’ with four or more impairments leading to a global rating of cognitive impairment. Following this definition, more than a fourth of our sample fulfilled the criterion of global impairment at both times. This is in line with findings of Andrykowski et al.,6 Meyers et al.2 and Booth-Jones et al.,27 as these results point to the fact that cognitive impairment in the course of chemotherapy may be conspicuous in a substantial subgroup of patients.

As the acute physical and psychosocial state of patients may influence cognitive performance, we looked for associations with medical and quality-of-life data. Concerning the extent of pretreatment, no differences between treatment intensities at either time could be detected. While a third of patients developed at least acute grade II GvHD, no association with neuropsychological results could be proven. Self-rated cognitive function was – corresponding with the literature – not correlated to test results. Interestingly, some of the symptom scales of the EORTC QLQ-C30 – shortness of breath, nausea and financial difficulties – were significantly related to cognitive performance at both times. Subjective complaints of shortness of breath and nausea may directly influence performance in memory and attention tests. However, the significant negative association between ability to produce words in a limited time and subjective perception of economic burden is difficult to interpret.

Since the work of van Dam et al.,28 in several studies and even in the public discussion, the ‘dose hypothesis’ expresses the observation that high-dose chemotherapies may lead to greater extents of cognitive deterioration than do standard therapies. This would support the expectation that SCT patient's cognitive performance will suffer due to the higher dose-conditioning regimen. The van Dam results were found retrospectively – later follow up could not replicate the finding.29

Another recent study comparing high-dose with standard-dose chemotherapy in breast cancer patients did not show a dose effect30 – moreover, there was a trend for pronounced neuropsychological deficits in the standard therapy group. The authors hypothesize that certain drugs such as methotrexate may contribute more to neuropsychological abnormalities. While these studies were conducted in breast cancer subjects, a recent cross-sectional analysis8 compared high-dose with standard treatment in hematology patients. Again, no significant differences between groups regarding cognitive performance were found regardless of dose.

Our findings are in contrast with the results of three prospective studies undertaken in hematology patients,2, 11, 12 reporting a significant decline in performance from admission to post transplant. While global impairment ratings were not performed in all studies, rates of impairment seemed to be higher than in our study. Differences between observed impairments are difficult to interpret, because cutoff scores for impairment vary largely between studies (1, 1.5 or 2 s.d.). Inspection of mean test scores of our patients shows that many tests declined between T0 and T1, but that the extent of this decline was – despite the potential for a practice effect – generally small. Even with less conservative cutoffs, we would not have detected significant impairments in our sample between assessments.

Another problem in comparing results is the assignment of neuropsychological test parameters to domains and subdomains of cognitive functions – this classification also varies largely between studies. Furthermore, there are the following important differences: in all aforementioned studies more ‘classical’ conditioning regimens, containing in most cases TBI and full dose of chemotherapy, were used. Our study is, to our knowledge, the first to compare standard and reduced-intensity conditioning. However, even between these two groups, significant differences are generally absent apart from one reasoning task, which is particularly susceptible to training effects. As TBI is known to be one potential factor predicting early cognitive deterioration4 and only one patient received TBI of 12 Gy in our sample, this might have further contributed to the lack of decline.

Although Sostak et al.12 observed a high prevalence of cognitive sequalae after transplant, 53% of their subjects with cognitive deficits before SCT had a ‘definitely favorable outcome after transplantation’ (p 844). This observation supports the hypothesis that there are subgroups of patients recovering fast in the course of SCT and others developing cognitive problems de novo. Comparing intraclass changes of impairments between assessments, our results do not support this hypothesis.

Furthermore, in our study as well as in others assessing candidates for stem cell transplant, cognitive deviations from norms were found even before the onset of chemotherapy treatment.2, 6, 8 This observation seems to be independent of pretreatment or quality-of-life aspects.8 On the other hand, the influence of emotional variables such as anxiety and depression on the development of cognitive performance was observed by Meyers et al.2 Consequently, we assumed that the sheer influence of receiving a transplant or high doses of systemic treatment is not sufficient to cause a decline of cognitive function. Additional factors including certain drugs with a pronounced effect on brain function such as methotrexate, additional drugs including corticosteroids and CSA, and psychological factors such as anxiety and depression before the admission for long inpatient treatment may add to problems in cognitive performance.

To date, the biomedical and psychosocial factors contributing to cognitive impairment in a subgroup of cancer patients in the course of, and following systemic treatment, are still largely unknown. Among others, cortical damage caused by certain drugs, cumulative effects of repeated treatments, psychological stressors or genetic risk factors31 are all relevant. However, mechanisms including the promotion of neurogenesis by human stem cells with a possible compensation of neuronal loss are in debate.32 Among other things, such mechanisms could have contributed to recovery of function during treatment and transplant.

There are several limitations of the study that have to be considered. First of all the sample size is rather small, and selective sampling cannot be excluded. It is possible that patients who could not be included into the study because of organizational or health-related reasons also had more problems in the cognitive domain. Comparing performance on admission against population norms means that nothing is known regarding premorbid levels of subjects before diagnosis. It is therefore possible that the incidence of impairment on admission may be underestimated. Although the test battery was composed carefully using parallel versions, training effects from first to second assessment cannot be excluded. In particular, tests examining executive functions are susceptible to these effects. Finally, it may be that the ‘true’ performance of patients at T0 is underestimated because of confounding psychological factors such as treatment-related distress, which may be especially pronounced on admission for allogeneic SCT. Assessment of adequate comparison groups as in the study of Harder et al.9 may be helpful with regard to this matter. Because the assessment of a variety of tests parallels multiple comparisons with norms and between times, the possibility of false significant results increases. We attempted to avoid this mistake by choosing a conservative level of significance of P0.01 for all analyses. We therefore consider that the significant deterioration in simple reaction time from T0 to T1 deserves particular notice. Obviously, this basic attention parameter is sufficiently sensitive to detect subtle impairment during the acute course of SCT.

Our results have important implications for health professionals and for future research. Our finding that a significant number of patients already performed significantly worse than the normal population on admission for SCT suggests that as well as the patients themselves, the professionals should be particularly aware of cognitive problems in this situation. Moreover, this result endorses neuropsychological assessment on a regular basis. In particular, patients who have to face intellectually demanding tasks in their working life should be carefully monitored. It is important to consider that, by the use of conservative cutoff scores, impairments in these patients might remain unidentified due to a high level of premorbid cognitive functioning. As training programs during acute follow-up and rehabilitation are available, staff should be aware of these possibilities.

Because of the multifaceted nature of the mechanisms involved and the high importance of such problems for individuals affected, further research using prospective designs with sufficient sample sizes and matched control groups should be undertaken.


  1. 1

    Ahles T, Tope D, Furstenberg C, Herndon J, Maurer LH, Cornblith AB et al. Psychologic and neuropsychologic impact of autologous bone marrow transplantation. J Clin Oncol 1996; 14: 1457–1462.

    CAS  Article  Google Scholar 

  2. 2

    Meyers CA, Weitzner M, Byrne K, Valentine A, Champlin RE, Przepiorka D . Evaluation of the neurobehavioral functioning of patients before, during, and after bone marrow transplantation. J Clin Oncol 1994; 12: 820–826.

    CAS  Article  Google Scholar 

  3. 3

    Wefel JS, Kayl AE, Meyers CA . Neuropsychological dysfunction associated with cancer and cancer therapies: a conceptual review of an emerging target. Br J Cancer 2004; 90: 1691–1696.

    CAS  Article  Google Scholar 

  4. 4

    Andrykowski MA, Altmaier EM, Barnett RL, Burish TG, Gingrich R, Henslee-Downey PJ . Cognitive dysfunction in adult survivors of allogeneic marrow transplantation: relationship to dose of total body irradiation. Bone Marrow Transplant 1990; 6: 269–276.

    CAS  PubMed  Google Scholar 

  5. 5

    Hjermstad MJ, Evensen SA, Kvaloy SO, Fayers PM, Kaasa S . Health-related quality of life 1 year after allogeneic or autologous stem-cell transplantation: a prospective study. J Clin Oncol 1999; 17: 706–718.

    CAS  Article  Google Scholar 

  6. 6

    Andrykowski MA, Schmitt FA, Gregg ME, Brady MJ, Lamb DG, Henslee-Downey PJ . Neuropsychologic impairment in adult bone marrow transplant candidates. Cancer 1992; 70: 2288–2297.

    CAS  Article  Google Scholar 

  7. 7

    Wefel JS, Lenzi R, Theriault R, Buzdar AU, Cruickshank S, Meyers CA . ‘Chemobrain’ in breast carcinoma?: a prologue. Cancer 2004; 101: 466–475.

    Article  Google Scholar 

  8. 8

    Harder H, Van Gool AR, Cornelissen JJ, Duivenvoorden HJ, Eijkenboom WM, Barge RM et al. Assessment of pre-treatment cognitive performance in adult bone marrow or haematopoietic stem cell transplantation patients: a comparative study. Eur J Cancer 2005; 41: 1007–1016.

    Article  Google Scholar 

  9. 9

    Harder H, Cornelissen JJ, Van Gool AR, Duivenvoorden HJ, Eijkenboom WM, van den Bent MJ . Cognitive functioning and quality of life in long-term adult survivors of bone marrow transplantation. Cancer 2002; 95: 183–192.

    Article  Google Scholar 

  10. 10

    Padovan CS, Yousry TA, Schleuning M, Holler E, Kolb HJ, Straube A . Neurological and neuroradiological findings in long-term survivors of allogeneic bone marrow transplantation. Ann Neurol 1998; 43: 627–633.

    CAS  Article  Google Scholar 

  11. 11

    Sostak P, Padovan CS, Yousry TA, Ledderose G, Kolb HJ, Straube A . Prospective evaluation of neurological complications after allogeneic bone marrow transplantation. Neurology 2003; 60: 842–848.

    CAS  Article  Google Scholar 

  12. 12

    Syrjala KL, Dikmen S, Langer S, Ruth-Roemer S, Abrams JR . Neuropsychological changes from pretransplant to one year in patients receiving myeloablative allogeneic hematopoietic cell transplant. Blood 2004; 104: 3386–3392.

    CAS  Article  Google Scholar 

  13. 13

    Anderson-Hanley C, Sherman ML, Riggs R, Agocha VB, Compas BE . Neuropsychological effects of treatment for adults with cancer: a meta-analysis and review of the literature. J Int Neuropsychol Soc 2003; 9: 967–982.

    Article  Google Scholar 

  14. 14

    Jansen CE, Miaskowski C, Dodd M, Dowling G, Kramer J . A metaanalysis of studies of the effects of cancer chemotherapy on various domains of cognitive function. Cancer 2005; 104: 2222–2233.

    Article  Google Scholar 

  15. 15

    Reitan RM . Trail Making Test: Manual for Administration and Scoring. Reitan Neuropsychology Laboratory: Tucson, AZ, 1979.

    Google Scholar 

  16. 16

    Zimmermann P, Fimm B . TAP Testbatterie zur Aufmerksamkeitsprüfung, Version 1.7. Psytest: Herzogenrath, 2002.

    Google Scholar 

  17. 17

    Brickenkamp R . Test d2 Aufmerksamkeits-Belastungs-Test, 9. Auflage. Hogrefe: Göttingen, 2002.

    Google Scholar 

  18. 18

    Wechsler D . WMS-R: Wechsler Memory Scale – Revised (Manual). The Psychological Corporation: San Antonio, 1987.

    Google Scholar 

  19. 19

    Härting C, Markowitsch HJ, Neufeld H, Calabrese P, Deisinger K (Hrsg.). WMS-R Wechsler Gedächtnistest – revidierte Fassung. Deutsche Adaptation der revidierten Fassung der Wechsler Memory Scale. Huber: Bern, 2000.

    Google Scholar 

  20. 20

    Rey A . L'examen clinique en psychologie. Presses Universitaires de France: Paris, 1964.

    Google Scholar 

  21. 21

    Helmstaedter C, Lendt M, Lux S . VLMT Verbaler Lern- und Merkfähigkeitstest. Beltz Test: Göttingen, 2001.

    Google Scholar 

  22. 22

    Aschenbrenner S, Tucha O, Lange KW . RWT Regensburger Wortflüssigkeits-Test. Hogrefe: Göttingen, 2000.

    Google Scholar 

  23. 23

    Horn W . L-P-S Leistungsprüfsystem, 2 Auflage. Göttingen: Hogrefe, 1983.

    Google Scholar 

  24. 24

    Aaronson NK, Ahmadzai S, Bergman B, Bullinger M, Cull A, Duez NJ 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.

    CAS  Article  Google Scholar 

  25. 25

    Spreen O, Strauss E . A Compendium of Neuropsychological Tests: Administration, Norms, and Commentary, 2nd edn. Oxford University Press: New York, NY, 1998.

    Google Scholar 

  26. 26

    Glucksberg H, Storb R, Fefer A, Buckner CD, Neimann PE, Clift RA et al. Clinical manifestations of graft-versus-host disease in human recipients of marrow from HLA-matched sibling donors. Transplantation 1974; 18: 295–304.

    CAS  Article  Google Scholar 

  27. 27

    Booth-Jones M, Jacobsen PB, Ransom S, Soety E . Characteristics and correlates of cognitive functioning following bone marrow transplantation. Bone Marrow Transplant 2005; 36: 695–702.

    CAS  Article  Google Scholar 

  28. 28

    van Dam FSAM, Schagen SB, Muller MJ, Boogert W, vd Wall E, Droogleever Fortuyn ME 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.

    CAS  Article  Google Scholar 

  29. 29

    Schagen SB, Muller MJ, Boogerd W, Rosenbrand RM, van Rijn D, Rodenhuis S et al. Late effects of adjuvant chemotherapy on cognitive function: a follow-up study in breast cancer patients. Ann Oncol 2002; 13: 1387–1397.

    CAS  Article  Google Scholar 

  30. 30

    Scherwath A, Mehnert A, Schleimer B, Schirmer L, Fehlauer F, Kreienberg R et al. Neuropsychological function on high-risk breast cancer survivors after stem-cell supported high-dose therapy versus standard-dose chemotherapy: evaluation of long-term treatment effects. Ann Oncol 2006; 17: 415–423.

    CAS  Article  Google Scholar 

  31. 31

    Ahles T, Saykin AJ, Noll WW, Furstenberg CT, Guerin S, Cole B et al. The relationship of APOE genotype to neuropsychological performance in long-term cancer survivors treated with standard dose chemotherapy. Psychooncology 2003; 12: 612–619.

    Article  Google Scholar 

  32. 32

    Altundag K, Moussallem CD, Baptista MZ . Promotion of neurogenesis by human stem cells in high-risk breast cancer survivals after stem-cell supported high-dose therapy. Ann Oncol 2006; 17: 1465.

    CAS  Article  Google Scholar 

Download references


This research was funded by the German José Carreras Leukemia Foundation e.V.

Author information



Corresponding author

Correspondence to F Schulz-Kindermann.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Schulz-Kindermann, F., Mehnert, A., Scherwath, A. et al. Cognitive function in the acute course of allogeneic hematopoietic stem cell transplantation for hematological malignancies. Bone Marrow Transplant 39, 789–799 (2007).

Download citation


  • stem cell transplantation
  • hematological malignancies
  • neuropsychological assessment
  • cognitive impairment
  • health-related quality of life

Further reading


Quick links