Journal of the Neurological Sciences 336 (2014) 132–137
Contents lists available at ScienceDirect
Journal of the Neurological Sciences journal homepage: www.elsevier.com/locate/jns
Cognitive fatigue in individuals with multiple sclerosis undergoing immunoablative therapy and hematopoietic stem cell transplantation Jason A. Berard a,b,⁎, Marjorie Bowman b, Harold L. Atkins b,c, Mark S. Freedman b,c, Lisa A.S. Walker a,b,c,d a
University of Ottawa, School of Psychology, Ottawa, Canada The Ottawa Hospital Research Institute, Ottawa, Canada University of Ottawa, Faculty of Medicine, Ottawa, Canada d Neuropsychology Service, The Ottawa Hospital, Ottawa, Canada b c
a r t i c l e
i n f o
Article history: Received 1 August 2013 Received in revised form 23 September 2013 Accepted 15 October 2013 Available online 23 October 2013 Keywords: Multiple sclerosis Cognitive fatigue PASAT Neuropsychology HSCT Fatigue
a b s t r a c t Background: Fatigue presents as a significant problem in multiple sclerosis (MS). Cognitive fatigue (CF) can be defined as a decrease in, or inability to maintain task performance throughout the duration of a continuous cognitive task. CF was evaluated using the Paced Auditory Serial Addition Test (PASAT) both pre- and postimmunoablation and hematopoietic stem cell transplantation (IA-HSCT) over a 3-year follow-up period. The magnitude of CF was examined and the impact of scoring methodology was evaluated. Methods: Twenty-three individuals with rapidly progressive MS and poor prognosis underwent high dose immunosuppression and subsequent HSCT. Individuals completed the 3″ and 2″ PASAT at baseline and every 6 months thereafter over a period of 36 months. As scoring methodology can impact its sensitivity to CF, the PASAT was scored according to three scoring methods. Results: CF was noted across all three scoring methods at baseline and at the majority of time points post-IA-HSCT on both the 3″ and 2″ PASAT. The magnitude of CF remained consistent both pre-and post-IA-HSCT. Conclusions: While results suggest that the procedure itself does not ameliorate an individual's susceptibility to CF; neither does it seem to negatively impact levels of CF. As such, results support the notion that the IA-HSCT procedure, despite its aggressive nature, does not exacerbate CF in this particular sample. © 2013 Elsevier B.V. All rights reserved.
1. Introduction Fatigue presents as a significant problem in multiple sclerosis (MS), occurring in up to 90% of those diagnosed [1]. Individuals with MS frequently report higher levels of fatigue when compared to healthy controls with the severity of fatigue being found to independently predict quality of life [2]. Despite the large body of literature examining MS-related fatigue, the concept of fatigue remains a poorly understood concept likely due to its multifaceted nature. In the past, research has focused predominantly on the study of physical fatigue; however, cognitive fatigue can often be equally as debilitating for those individuals affected. Cognitive fatigue (CF) can be defined as a decrease in, or inability to sustain, task performance throughout the duration of a continuous information processing speed (IPS) task [3,4]. During IPS tasks, individuals with MS show more susceptibility to the effects of CF when compared to healthy controls [4]. Individuals susceptible to CF are less able to maintain the required cognitive effort necessary to
⁎ Corresponding author at: The Ottawa Hospital Research Institute, 501 Smyth Road, Suite 7300, Ottawa, Ontario, K1H 8L6, Canada. Tel.: +1 613 737 8899x73875; fax: +1 613 737 8895. E-mail address: [email protected] (J.A. Berard). 0022-510X/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.jns.2013.10.023
continuously meet task demands on an IPS task over time. This is often reflected by a breakdown in task performance as the task progresses. Previous work by our group examining IPS tasks as potential measures of CF has shown that, despite its limitations [5], the Paced Auditory Serial Addition Test (PASAT) can serve as a reliable and sensitive measure of CF in MS [6]. 1.1. PASAT scoring methods The PASAT is a measure of IPS and working memory (WM) in which participants are presented with a series of single digit numbers (60 trials) where the two most recent digits heard must be summed out loud. The speed at which information must be processed can be manipulated by altering the interstimulus intervals (ISI). CF can be measured using the PASAT by comparing performance on the first half of the task (i.e. the first 30 trials) versus the second half of the task (i.e. the last 30 trials). Despite the challenges associated with the use of the PASAT [5–10], it remains one of the most sensitive measures of IPS and WM deficits in MS [10,11]; and as such has been continuously used in studies examining CF in MS where sustained cognitive effort is required [4,6,12]. Traditionally, PASAT performance is scored by counting the overall number of correct responses. Individuals may, however, adopt a “chunking method” strategy such that the first two numbers are added, the next number is skipped, the following two numbers are
J.A. Berard et al. / Journal of the Neurological Sciences 336 (2014) 132–137
added, and so forth [10]. This strategy reduces the overall difficulty of the task by decreasing the need to perform multiple cognitive processes simultaneously (i.e. diminishes the working memory demands). While these individuals may still achieve a score within normal limits, their performance no longer reflects the ability to meet the task demands as intended. The number of correct dyad responses better reflects an individual's ability to successfully meet task demands [6,10]. In this case, a correct score is only assigned when one correct response immediately precedes another; thus, dyad scoring provides an indication of performance when executing the task as intended. As such, the number of correct responses and the number of correct dyads both provide a measure of performance level. One may further calculate a percent dyad score; an indication of the proportion of time an individual is performing the task as intended. While not a measure of performance accuracy per se, higher percent dyad scores reflect a greater ability to produce correct responses in accordance with the task demands [10]. The percent dyad scoring method thus provides a reflection of an individual's performance strategy.
1.2. IA-HSCT procedure Presently, no curative treatment exists for MS. While current treatment therapies aim to slow the progression of the disease, results have been limited in their success. In an attempt to eradicate the autoimmune processes which give rise to MS, the Ottawa Hospital MS Clinic (as well as other groups worldwide) has adopted a Bone Marrow Transplant approach to treatment in hopes of establishing long-lasting periods during which progression of the disease is halted [13]. Given the intensive nature of the procedure, only a select group of individuals with rapidly-progressing MS and poor prognosis was chosen to undergo immunoablative therapy and hematopoietic stem cell transplantation (IA-HSCT). Of the 23 individuals to undergo the procedure, no new attacks or MRI lesions have been reported [14, Freedman & Atkins, 2013 personal communication]. Neuroimaging has revealed, however, that the procedure itself results in a rate of brain atrophy significantly greater than is expected on the basis of natural disease progression. Individuals experienced a 3.2% (median) decrease in total brain volume over 2.4 (median) months [15]. Preliminary results suggest that atrophy may not have any lasting negative impact on cognition [16], but to date, the potential impact on cognitive fatigue has not been addressed. Given the aggressive nature of the treatment alongside the increased rate of atrophy noted, the current study aims to evaluate levels of CF in the 23 individuals before and after undergoing the IA-HSCT procedure. To the best of our knowledge, no research to date has evaluated the impact of such a procedure on levels of CF in an MS population.
1.3. Hypotheses The primary objective was to examine performance across the task on the 3″ and 2″ PASAT in the 23 individuals both pre- and post-IAHSCT in order to determine whether the procedure influenced levels of CF. Furthermore, we determined whether PASAT scoring method influenced its sensitivity in detecting CF in this particular sample. It was hypothesized that CF would be evident both pre-and post IAHSCT as evidenced by lower performance on the 2nd half of the PASAT when compared to the 1st half of the PASAT at all time points. Furthermore, given the aggressive nature of the procedure, we hypothesized that the magnitude of CF (i.e. the degree of change between 1st and 2nd half performance) would be greater for those time points immediately following the IA-HSCT procedure when compared to the magnitude of CF at baseline; with the degree of change returning to baseline levels with temporal distance from the procedure.
133
2. Methods 2.1. Participants This study was approved by the Ottawa Hospital Research Ethics Board and informed consent was obtained. A total of thirty-four individuals were screened as potential candidates for the study. Eleven individuals did not meet specific inclusion criteria; as such, twentythree individuals (14 females; 9 males) with rapidly progressing MS who failed to respond to routine therapy were enrolled. High risk of progression was defined as ≥5 relapses in the first two years of the disease or attainment of a Functional System (FS) Score of at least 3 affecting pyramidal/cerebellar subscores within 5 years of disease onset. If a patient had previously received a cytotoxic agent (Mitoxantrone or Cyclophosphamide) they must have had normal bone marrow morphology and cytogenetics before being considered eligible. Age ranged from 23 to 44 years (mean = 32.65 (5.82) years) and education ranged from completion of high school to completion of graduate degrees. Expanded Disability Status Scale (EDSS) scores pre-IA-HSCT ranged from 1.5 to 6.5 (mean = 4.87 (1.40)). Twelve individuals were diagnosed with relapsing–remitting MS (RRMS) with the remaining 11 diagnosed with secondary progressive MS (SPMS). Those with primary progressive MS were excluded. 2.2. Procedure The study was a tri-center phase II efficacy study of the role of IAHSCT on the natural history of MS. Participants underwent stem cell mobilization with IV cyclophosphamide (4.5 g/m2) and 10 days of granulocyte colony-stimulating factor (10 μg/kg/day) followed by stem cell collection using peripheral vein leukapheresis. All stem cell grafts were CD34 selected and cryopreserved until transplantation. Immunoablation was accomplished in the first six individuals using cyclophosphamide (200mg/kg), oral bulsufan (16 mg/kg), and IV rabbit antithymocyte globulin (5 mg/kg). A similar regimen was used for the remaining 17 individuals although these individuals received doseadjusted IV busulfan (9.6 mg/kg) in place of the oral bulsufan. Steroids were administered concurrently with chemotherapy. Patients did not receive further MS-disease modifying drugs or experimental therapy after HSCT. The PASAT was administered by a trained research assistant. Participants completed the 3″ and 2″ PASAT at baseline (pre-IA-HSCT) and serially every 6 months post-IA-HSCT for a period of 3 years (i.e. 36 months). PASAT performance was compared between the first and second half of the task at all time points for both the 3″ and 2″ ISIs. Responses were recorded and the following scores were obtained: total number of correct responses, total correct dyad score, and percent dyad score. According to convention, in order to have an equal number of dyad responses in each half (i.e. 30 possible dyads in both halves), a dyad score was given for a correct response to the first pair of numbers presented; despite there being no possible preceding response. Percent dyad scores were obtained using the following formula: (1− ((total correct score-dyad score)/total correct score)) × 100% [6,10]. Two participants did not complete follow-up evaluations past 18 months as they did not comply with the study demands and inclusion of their data would have made it more difficult to interpret findings in the context of the treatment of interest. 2.3. Data analysis In order to examine performance differences between the first and second half of the PASAT, data was analysed for both the 3″ and 2″ ISIs using paired-samples t-tests. This was repeated for all three scoring methods and across all time points. In order to determine the magnitude of CF at each time, difference scores were calculated such that scores from the first half of the task were subtracted from scores
134
J.A. Berard et al. / Journal of the Neurological Sciences 336 (2014) 132–137
at the second half in order to create a half-change score. In order to examine whether the magnitude of CF differed from baseline following the IA-HSCT procedure, paired-samples t-tests were used which compared half-change scores at baseline with half-change scores at all times post-IA-HSCT. This was repeated for all three scoring methods and for both the 3″ and 2″ ISIs. It should be highlighted that given the small sample size and the limited power, more sophisticated statistical analyses were not possible.
3. Results 3.1. Cognitive fatigue and scoring method sensitivity Evidence of cognitive fatigue was noted across all three scoring methods on both the 3″ and 2″ ISIs. This was reflected by a significant breakdown in task performance on the second half of the task when compared to the first half of the task (Figs. 1 & 2). Cognitive fatigue was noted at the majority of time points both pre-and post-IA-HSCT with the exception of the 3″ ISI at the 6 month follow-up and the 2″
Fig. 2. Mean 2″ PASAT scores across scoring method.
ISI at the 30 month follow-up when the total number of correct responses was examined (Table 1). Given that cognitive fatigue was noted across all three scoring methods, the way in which the PASAT is scored does not seem to influence its sensitivity in detecting CF in this particular sample. Overall, however, the dyad and percent dyad scores were more consistent in their sensitivity to CF over time as CF was noted at all time points using these two scoring methods. 3.2. Magnitude of cognitive fatigue over time Paired samples t-test analyses (Table 2) showed no significant differences in the degree of cognitive fatigue from baseline to any follow-up session post-IA-HSCT. That is, the extent to which performance declined between the first and second half of the task was comparable both before and after the procedure. These results were consistent for both the 3″ and 2″ ISIs and across all three scoring methods. 4. Discussion
Fig. 1. Mean 3″ PASAT scores across scoring method.
Consistent with previous research examining the PASAT as a measure of cognitive fatigue [4,6,10], evidence of CF was noted in our sample.
J.A. Berard et al. / Journal of the Neurological Sciences 336 (2014) 132–137
135
Table 1 Mean first and second half total correct, correct dyad, and percent dyad scores at each administration. Mean total correct (SD)
p
Mean correct dyad (SD)
p
Mean percent dyad (SD)
p
Baseline 3″ ISI – First Half 3″ ISI – Second Half 2″ ISI – First Half 2″ ISI – Second Half
21.52 (5.40) 19.78 (6.19) 17.27 (5.80) 15.14 (4.99)
.044
16.74 (7.45) 14.09 (7.92) 11.91 (6.39) 8.36 (6.09)
.023
73.25 (20.86) 64.61 (24.00) 63.65 (18.44) 48.90 (23.67)
.020
6 Month 3″ ISI – First Half 3″ ISI – Second Half 2″ ISI – First Half 2″ ISI – Second Half
23.00 (6.13) 22.50 (5.52) 17.14 (6.42) 15.27 (5.82)
.460
19.23 (8.18) 16.82 (7.80) 11.77 (8.06) 8.64 (6.88)
.045
78.10 (22.09) 69.82 (22.45) 60.60 (24.76) 48.72 (25.26)
.007
12 Month 3″ ISI – First Half 3″ ISI – Second Half 2″ ISI – First Half 2″ ISI – Second Half
23.50 (6.15) 21.14 (7.04) 18.59 (5.99) 14.82 (6.28)
.006
20.00 (7.84) 16.00 (9.73) 13.00 (7.26) 8.73 (6.96)
.006
80.51 (19.92) 68.53 (24.05) 64.24 (19.79) 50.08 (24.83)
.008
18 Month 3″ ISI – First Half 3″ ISI – Second Half 2″ ISI – First Half 2″ ISI – Second Half
24.00 (5.77) 21.32 (6.62) 18.62 (6.55) 17.00 (5.44)
.001
20.68 (7.69) 16.32 (8.99) 14.05 (8.03) 10.95 (6.83)
.000
82.47 (17.25) 68.76 (26.77) 69.24 (22.02) 59.45 (21.95)
.001
24 Month 3″ ISI – First Half 3″ ISI – Second Half 2″ ISI – First Half 2″ ISI – Second Half
24.33 (5.37) 22.83 (6.30) 18.67 (7.10) 16.17 (7.10)
.019
21.28 (7.41) 18.50 (8.55) 14.33 (8.60) 10.50 (8.20)
.004
84.68 (13.69) 75.19 (22.65) 70.14 (20.98) 53.83 (28.63)
.006
30 Month 3″ ISI – First Half 3″ ISI – Second Half 2″ ISI – First Half 2″ ISI – Second Half
25.79 (4.44) 21.57 (7.61) 17.79 (5.63) 15.57 (6.81)
.019
36 Month 3″ ISI – First Half 3″ ISI – Second Half 2″ ISI – First Half 2″ ISI – Second Half
26.19 (4.89) 23.38 (7.01) 20.20 (4.72) 17.00 (5.52)
.019
.035
.036
.000
.038
.000
.102
.003
.003
.002
.000
.001
.000
23.14 (6.51) 17.50 (9.83) 12.57 (6.72) 9.36 (7.58)
.005
23.88 (7.34) 19.19 (10.80) 15.27 (6.62) 10.73 (6.79)
.014
These results support the notion that the PASAT is a reliable and sensitive measure of CF in MS [10,11]. A significant breakdown in task performance was noted on the second half of the task when compared to the first half; indicative of an inability of our sample to maintain the required cognitive effort necessary to continuously and successfully meet the task demands over time (Figs. 1 & 2). These results were consistent at both the 3″ and 2″ ISIs; thus CF was apparent regardless of overall task complexity. Previous work which examined the relationship between task complexity and levels of CF in a sample of early-phase RRMS found that, in contrast to results from the current study, CF was not apparent on the 2″ ISI [6]. It was suggested that the lack of CF noted at this more difficult ISI may have been due to overall levels of poor performance across the task (i.e. there was no breakdown in performance noted over time as performance was already quite low to begin with). In contrast, the current sample showed a trend for greater initial performance on the first half of the task when compared to the previous study and as such, a measureable breakdown in performance (i.e. CF) occurred between the first and second half of the task. The inconsistency between these two results may be due to differences in the disease characteristics of the individuals comprising the two samples given that our previous study focused on a homogenous sample of early-phase RRMS, whereas the current study focused on individuals with rapidly progressive RRMS or SPMS with poor prognosis. One can speculate that the vulnerability to CF noted on the 2″ ISI in the current sample may be attributed to the differences in disease subtypes or disease durations as compared to our previous study. To date, however, little research has examined differences in susceptibility to CF between individuals with differing subtypes of MS, as well as between those with longer versus shorter disease durations.
.033
.002
.000
.003
.000
.004
.000
87.53 (14.20) 73.74 (24.78) 65.14 (20.58) 50.04 (27.77)
.006
87.77 (20.12) 73.49 (31.82) 72.65 (17.29) 57.09 (23.56)
.021
.006
.001
Given the small sample size in the current study, it was not possible to examine differences in disease characteristics in this particular sample. Our conclusions, therefore, are purely speculation. As such, future studies should examine the impact of disease subtype (as well as in those who present with clinically isolated syndrome) and disease duration on levels of CF in order to determine whether scoring methodology impacts PASAT sensitivity to CF between these groups. Previous research has shown that the percent dyad scoring method was the most sensitive measure of CF on the PASAT [6]. Results from the current study, however, showed no greater sensitivity with one scoring method over the others. Although the dyad and percent dyad scores yielded more consistent findings across time, cognitive fatigue was noted across all three scoring methods; suggesting that, unlike previous work, in this sample a measure of performance strategy (i.e. percent dyad scores) does not appear to be more sensitive to CF when compared to measures of performance level (i.e. total number correct and total correct dyads). We can speculate that this inconsistency between the two studies may again be attributed to differences in disease characteristics between the two samples as discussed above. Cognitive fatigue was noted at all time points both pre- and post-IAHSCT, with the exception of the 3″ ISI at the 6month follow-up and the 2″ ISI at the 30 month follow-up when the total number of correct responses was examined (Table 1). While these two anomalous findings suggest that performance remained relatively stable throughout the task at these times (i.e. there was a lack of CF), this anomaly is unlikely to be clinically meaningful given the overall pattern for these individuals to exhibit cognitive fatigue at all other times, as well as the findings supporting CF at these time points when using the other scoring methods. Generally, individuals in this sample were vulnerable to CF
136
J.A. Berard et al. / Journal of the Neurological Sciences 336 (2014) 132–137
Table 2 Degree of change in cognitive fatigue from baseline to all follow-up sessions post-IA-HSCT. Mean total correct (SD)
p
Mean correct dyad (SD)
p
Mean percent dyad (SD)
p
Baseline to 6 Months 3″ ISI – Half-Change @ Baseline 3″ ISI – Half-Change @ 6 Months 2″ ISI – Half-Change @ Baseline 2″ ISI – Half-Change @ 6 Months
1.77 (3.99) .50 (3.11) 2.14 (4.43) 1.86 (3.91)
.172
2.77 (5.30) 2.41 (5.29) 3.55 (4.88) 3.14 (4.27)
.787
9.09 (16.70) 8.28 (13.03) 14.75 (15.52) 11.88 (16.36)
.839
Baseline to 12 Months 3″ ISI – Half-Change @ Baseline 3″ ISI – Half-Change @ 12 Months 2″ ISI – Half-Change @ Baseline 2″ ISI – Half-Change @ 12 Months
1.68 (3.98) 2.36 (3.66) 1.95 (4.46) 3.95 (3.31)
.538
2.50 (5.28) 4.00 (6.07) 3.38 (4.93) 4.48 (4.02)
.387
Baseline to 18 Months 3″ ISI – Half-Change @ Baseline 3″ ISI – Half-Change @ 18 Months 2″ ISI – Half-Change @ Baseline 2″ ISI – Half-Change @ 18 Months
1.55 (3.88) 2.68 (3.40) 2.10 (4.54) 1.62 (3.34)
.341
2.27 (5.00) 4.36 (4.77) 3.62 (4.98) 3.10 (3.77)
.174
Baseline to 24 Months 3″ ISI – Half-Change @ Baseline 3″ ISI – Half-Change @ 24 Months 2″ ISI – Half-Change @ Baseline 2″ ISI – Half-Change @ 24 Months
.72 (3.43) 1.50 (2.46) 1.71 (4.62) 2.41 (2.29)
.454
1.33 (4.92) 2.78 (3.49) 3.18 (5.21) 3.94 (3.70)
.307
Baseline to 30 Months 3″ ISI – Half-Change @ Baseline 3″ ISI – Half-Change @ 30 Months 2″ ISI – Half-Change @ Baseline 2″ ISI – Half-Change @ 30 Months
1.07 (3.43) 4.21 (5.86) 2.31 (5.34) 2.00 (4.83)
.095
1.79 (5.04) 5.64 (6.25) 3.77 (5.91) 3.31 (5.25)
.120
Baseline to 36 Months 3″ ISI – Half-Change @ Baseline 3″ ISI – Half-Change @ 36 Months 2″ ISI – Half-Change @ Baseline 2″ ISI – Half-Change @ 36 Months
1.38 (3.42) 2.81 (4.26) 2.20 (4.96) 3.20 (3.41)
.845
.104
.701
.619
.890
.292 .531
both before and after the IA-HSCT procedure. It is of note, however, that the procedure itself did not lead to any greater vulnerability to CF given that individuals showed a similar magnitude of CF both pre-and post-IAHSCT (Table 2). That is, the degree in which task performance declined across the task was comparable at all time points, suggesting the procedure itself (or presumably the associated atrophy) does not seem to negatively impact an individual’s susceptibility to CF. As seen in Figs. 1 & 2, there was a trend for subjects to perform better on the PASAT with each successive administration; consistent with the practice effects that are usually seen with repeat testing [10,17,18]. As part of the larger overall project which includes the current study, we examined the impact of the IA-HSCT procedure on overall PASAT performance and found that after taking practice effects into account (using reliable change methodology) there still remained an overall trend for improvement in performance (while non-significant) across the follow-up sessions [19]. It is notable that despite this overall trend for increased performance, individuals' vulnerability to CF did not show a similar pattern of improvement over time. Individuals in this sample continued to remain vulnerable to CF across the follow-up sessions when performing a sustained and cognitively demanding task (as evidenced by a breakdown in their performance across the task) despite their overall increase in performance levels. At this time, it is difficult to say if the lack of impact of the IA-HSCT procedure on levels of CF would also be noted using traditional treatment therapies. While studies have examined the relationship between some traditional therapies and levels of overall fatigue in MS [20,21], few studies have examined the relationship between these treatments and CF specifically. As such, future studies should examine changes in CF following these traditional therapies in order to determine how they compare with those changes seen following the IA-HSCT procedure. Along with previous work by our group examining the impact of the IA-HSCT procedure on measures of cognition and neuroimaging [16,19],
2.25 (5.03) 4.69 (6.71) 3.60 (5.50) 4.53 (4.47)
.775
.391
.737
.659
.843
.207 .625
8.31 (16.76) 11.98 (19.22) 14.40 (15.82) 14.84 (13.73)
6.64 (13.67) 13.72 (15.92) 15.43 (15.58) 9.79 (13.78)
3.58 (12.51) 9.49 (12.81) 13.55 (14.95) 14.91 (11.72)
4.48 (13.79) 13.79 (15.72) 15.21 (16.43) 14.34 (17.79)
5.22 (13.18) 14.27 (22.18) 14.23 (15.44) 15.57 (14.16)
.515
.472 .921
.136 .304
.160 .767
.160 .896
.122 .799
the current study highlights the promise of the IA-HSCT procedure as a viable treatment option for individuals with MS who do not respond well to traditional treatment therapies given it seems the procedure itself does not further exacerbate CF. While results suggest that the procedure itself does not ameliorate an individual's susceptibility to CF; neither does it seem to negatively impact levels of CF despite its aggressive nature. One limitation of the current study however is the lack of a control group. While attempts were made to recruit a suitable control group comprising individuals undergoing bone marrow transplant for other indications besides MS, attempts were unsuccessful. Although we conclude that the IA-HSCT procedure had no impact on individuals' level of CF and that individuals remained vulnerable to CF despite a trend for overall improvement in performance, a control group would have allowed us to examine whether these results were specific to individuals with rapidly progressive RRMS or SPMS undergoing IA-HSCT. Overall, while there seems to be no detrimental impact on levels of CF for those individuals who have undergone IA-HSCT, CF does continue to present as an ongoing problem for these individuals post-procedure. Evidence from the current study suggests that, despite the apparent improvements in PASAT performance over time, individuals continue to show a breakdown in task performance indicative of CF as the task progresses. These findings have important clinical relevance for the study of CF when one considers that individuals may present with apparently preserved cognitive performance, but may subjectively report feeling as though they have to mentally work harder in order to achieve or maintain this level of performance. Functional neuroimaging studies would allow for the evaluation of changes in neural activity across the task, and may provide objective evidence to substantiate these subjective reports. While structural neuroimaging variables were collected for the current participants, functional neuroimaging was not performed. The fact that cognitive performance seems to improve (albeit to a non-significant degree) following IA-HSCT, but cognitive
J.A. Berard et al. / Journal of the Neurological Sciences 336 (2014) 132–137
fatigue does not, suggests that these two concepts are distinct, and provides confirmatory evidence for anecdotal reports from individuals with MS who say they can often maintain adequate levels of cognitive performance compared to their peers, but only at the cost of high levels of cognitive fatigue and mental effort. In conclusion, despite the fact that no improvements in cognitive fatigue were noted over time, the current findings provide important safety information about IA-HSCT, in that the procedure appears to have no detrimental impact on levels of cognitive fatigue, either immediately post-procedure, or up to three years post-procedure. This information, coupled with the findings of a potential mild benefit on cognitive performance, lends further credence to the use of IA-HSCT given that the procedure, despite its aggressive nature, does not seem to further exacerbate CF. Conflict of interest statement The authors declare that they have no conflict of interest. Acknowledgments We would like to gratefully thank the participants of this study for their time and effort. This research was supported by a grant from the Research Foundation of the Multiple Sclerosis Society of Canada. References [1] Kinkel RP. Fatigue in multiple sclerosis: reducing the impact through comprehensive management. Int J MS Care 2000;2(3):3–10. [2] Amato MP, Ponziani G, Rossi F, Liedl CL, Stefanile C, Rossi L. Quality of life in multiple sclerosis: the impact of depression, fatigue, and disability. Mult Scler 2001;7:340–4. [3] Schwid SR, Covington M, Segal BM, Goodman AD. Fatigue in multiple sclerosis: current understanding and future directions. J Rehabil Res Dev 2002;39:211–24. [4] Bryant M, Chiaravalloti ND, DeLuca J. Objective measurement of cognitive fatigue in multiple sclerosis. Rehabil Psychol 2004;49:114–22. [5] Tombaugh TN. A comprehensive review of the Paced Auditory Serial Addition Test. Arch Clin Neuropsychol 2006;21:53–76.
137
[6] Walker LAS, Berard JA, Berrigan LI, Rees LM, Freedman MS. Detecting cognitive fatigue in multiple sclerosis: method matters. J Neurol Sci 2012;316:86–92. [7] Walker LAS, Cheng A, Berard JA, Berrigan LI, Rees LM, Freedman MS. Tests of information processing speed – what do people with multiple sclerosis think about them? Int J MS Care 2012;14:92–9. [8] Sl Barker-Collo. Within session practice effects on the PASAT in clients with multiple sclerosis. Arch Clin Neuropsychol 2005;20:145–52. [9] Brittain JL, LaMarche JA, Reeder KP, Roth DL, Boll TJ. Effects of age and IQ on Paced Auditory Serial Addition Test (PASAT) performance. Clin Neuropsychol 1991;5:163–75. [10] Fisk JD, Archibald CJ. Limitations of the Paced Auditory Serial Addition Test (PASAT) as a measure of working memory in patients with multiple sclerosis. J Int Neuropsychol Soc 2001;7:363–72. [11] Forn C, Belenguer A, Parcet-Ibars MA, Avila C. Information-processing speed is the primary deficit underlying the poor performance of multiple sclerosis patients in the Paced Auditory Serial Addition Test (PASAT). J Clin Exp Neuropsychol 2008;30:789–96. [12] Schwid SR, Tyler CM, Scheid EA, Weinstein A, Goodman AD, McDermott MP. Cognitive fatigue during a test requiring sustained attention: a pilot study. Mult Scler 2003;9:503–8. [13] Pasquini MC, Griffith LM, Arnold DL, Atkins HL, Bowen JD, Chen JT, et al. Hematopoietic stem cell transplantation for multiple sclerosis: collaboration of the CIBMTR and EBMT to facilitate international clinical studies. Biol Blood Marrow Transplant 2010;16(8):1076–83. [14] Freedman MS. Bone marrow transplantation: does it stop MS progression? J Neurol Sci 2007;259:85–9. [15] Chen JT, Collins DL, Atkins HL, Freedman MS, Galai A, Arnold DL, et al. Brain atrophy after immunoablation and stem cell transplantation in multiple sclerosis. Neurology 2006;66:1935–7. [16] Walker LAS, Berard JA, Bowman M, Atkins HL, Lee H, Freedman MS. Cognitive change and neuroimaging following immunoablative therapy and hematopoietic stem cell transplantation in multiple sclerosis: a pilot study. Multi Scler Relat Disord 2013. http://dx.doi.org/10.1016/j.msard.2013.05.001. [17] Solari A, Radice D, Manneschi L, Motti L, Montanari E. The multiple sclerosis functional composite: different practice effects in the three test components. J Neurol Sci 2005;228:71–4. [18] McCaffrey RJ, Ortega A, Orsillo SM, Nelles WB, Haase RF. Practice effects in repeated neuropsychological assessments. Clin Neuropsychol 1992;6:32–42. [19] Walker LAS, Berard JA, AtkinsHL, Bowman M, LeeH, FreedmanMS. Longitudinal change in PASAT performance following immunoablative therapy and haematopoietic stem cell transplant in multiple sclerosis. [Manuscript in progress]. [20] Walther EV, Hohfeld R. Multiple sclerosis: side effects of interferon beta therapy and their management. Neurology 1999;10:1622–7. [21] Putzki N, Yaldizli Ö, Tettenborn B, Diener HC. Multiple sclerosis associated fatigue during natalizumab treatment. J Neurol Sci 2009;285:109–13.
Contents lists available at ScienceDirect
Journal of the Neurological Sciences journal homepage: www.elsevier.com/locate/jns
Cognitive fatigue in individuals with multiple sclerosis undergoing immunoablative therapy and hematopoietic stem cell transplantation Jason A. Berard a,b,⁎, Marjorie Bowman b, Harold L. Atkins b,c, Mark S. Freedman b,c, Lisa A.S. Walker a,b,c,d a
University of Ottawa, School of Psychology, Ottawa, Canada The Ottawa Hospital Research Institute, Ottawa, Canada University of Ottawa, Faculty of Medicine, Ottawa, Canada d Neuropsychology Service, The Ottawa Hospital, Ottawa, Canada b c
a r t i c l e
i n f o
Article history: Received 1 August 2013 Received in revised form 23 September 2013 Accepted 15 October 2013 Available online 23 October 2013 Keywords: Multiple sclerosis Cognitive fatigue PASAT Neuropsychology HSCT Fatigue
a b s t r a c t Background: Fatigue presents as a significant problem in multiple sclerosis (MS). Cognitive fatigue (CF) can be defined as a decrease in, or inability to maintain task performance throughout the duration of a continuous cognitive task. CF was evaluated using the Paced Auditory Serial Addition Test (PASAT) both pre- and postimmunoablation and hematopoietic stem cell transplantation (IA-HSCT) over a 3-year follow-up period. The magnitude of CF was examined and the impact of scoring methodology was evaluated. Methods: Twenty-three individuals with rapidly progressive MS and poor prognosis underwent high dose immunosuppression and subsequent HSCT. Individuals completed the 3″ and 2″ PASAT at baseline and every 6 months thereafter over a period of 36 months. As scoring methodology can impact its sensitivity to CF, the PASAT was scored according to three scoring methods. Results: CF was noted across all three scoring methods at baseline and at the majority of time points post-IA-HSCT on both the 3″ and 2″ PASAT. The magnitude of CF remained consistent both pre-and post-IA-HSCT. Conclusions: While results suggest that the procedure itself does not ameliorate an individual's susceptibility to CF; neither does it seem to negatively impact levels of CF. As such, results support the notion that the IA-HSCT procedure, despite its aggressive nature, does not exacerbate CF in this particular sample. © 2013 Elsevier B.V. All rights reserved.
1. Introduction Fatigue presents as a significant problem in multiple sclerosis (MS), occurring in up to 90% of those diagnosed [1]. Individuals with MS frequently report higher levels of fatigue when compared to healthy controls with the severity of fatigue being found to independently predict quality of life [2]. Despite the large body of literature examining MS-related fatigue, the concept of fatigue remains a poorly understood concept likely due to its multifaceted nature. In the past, research has focused predominantly on the study of physical fatigue; however, cognitive fatigue can often be equally as debilitating for those individuals affected. Cognitive fatigue (CF) can be defined as a decrease in, or inability to sustain, task performance throughout the duration of a continuous information processing speed (IPS) task [3,4]. During IPS tasks, individuals with MS show more susceptibility to the effects of CF when compared to healthy controls [4]. Individuals susceptible to CF are less able to maintain the required cognitive effort necessary to
⁎ Corresponding author at: The Ottawa Hospital Research Institute, 501 Smyth Road, Suite 7300, Ottawa, Ontario, K1H 8L6, Canada. Tel.: +1 613 737 8899x73875; fax: +1 613 737 8895. E-mail address: [email protected] (J.A. Berard). 0022-510X/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.jns.2013.10.023
continuously meet task demands on an IPS task over time. This is often reflected by a breakdown in task performance as the task progresses. Previous work by our group examining IPS tasks as potential measures of CF has shown that, despite its limitations [5], the Paced Auditory Serial Addition Test (PASAT) can serve as a reliable and sensitive measure of CF in MS [6]. 1.1. PASAT scoring methods The PASAT is a measure of IPS and working memory (WM) in which participants are presented with a series of single digit numbers (60 trials) where the two most recent digits heard must be summed out loud. The speed at which information must be processed can be manipulated by altering the interstimulus intervals (ISI). CF can be measured using the PASAT by comparing performance on the first half of the task (i.e. the first 30 trials) versus the second half of the task (i.e. the last 30 trials). Despite the challenges associated with the use of the PASAT [5–10], it remains one of the most sensitive measures of IPS and WM deficits in MS [10,11]; and as such has been continuously used in studies examining CF in MS where sustained cognitive effort is required [4,6,12]. Traditionally, PASAT performance is scored by counting the overall number of correct responses. Individuals may, however, adopt a “chunking method” strategy such that the first two numbers are added, the next number is skipped, the following two numbers are
J.A. Berard et al. / Journal of the Neurological Sciences 336 (2014) 132–137
added, and so forth [10]. This strategy reduces the overall difficulty of the task by decreasing the need to perform multiple cognitive processes simultaneously (i.e. diminishes the working memory demands). While these individuals may still achieve a score within normal limits, their performance no longer reflects the ability to meet the task demands as intended. The number of correct dyad responses better reflects an individual's ability to successfully meet task demands [6,10]. In this case, a correct score is only assigned when one correct response immediately precedes another; thus, dyad scoring provides an indication of performance when executing the task as intended. As such, the number of correct responses and the number of correct dyads both provide a measure of performance level. One may further calculate a percent dyad score; an indication of the proportion of time an individual is performing the task as intended. While not a measure of performance accuracy per se, higher percent dyad scores reflect a greater ability to produce correct responses in accordance with the task demands [10]. The percent dyad scoring method thus provides a reflection of an individual's performance strategy.
1.2. IA-HSCT procedure Presently, no curative treatment exists for MS. While current treatment therapies aim to slow the progression of the disease, results have been limited in their success. In an attempt to eradicate the autoimmune processes which give rise to MS, the Ottawa Hospital MS Clinic (as well as other groups worldwide) has adopted a Bone Marrow Transplant approach to treatment in hopes of establishing long-lasting periods during which progression of the disease is halted [13]. Given the intensive nature of the procedure, only a select group of individuals with rapidly-progressing MS and poor prognosis was chosen to undergo immunoablative therapy and hematopoietic stem cell transplantation (IA-HSCT). Of the 23 individuals to undergo the procedure, no new attacks or MRI lesions have been reported [14, Freedman & Atkins, 2013 personal communication]. Neuroimaging has revealed, however, that the procedure itself results in a rate of brain atrophy significantly greater than is expected on the basis of natural disease progression. Individuals experienced a 3.2% (median) decrease in total brain volume over 2.4 (median) months [15]. Preliminary results suggest that atrophy may not have any lasting negative impact on cognition [16], but to date, the potential impact on cognitive fatigue has not been addressed. Given the aggressive nature of the treatment alongside the increased rate of atrophy noted, the current study aims to evaluate levels of CF in the 23 individuals before and after undergoing the IA-HSCT procedure. To the best of our knowledge, no research to date has evaluated the impact of such a procedure on levels of CF in an MS population.
1.3. Hypotheses The primary objective was to examine performance across the task on the 3″ and 2″ PASAT in the 23 individuals both pre- and post-IAHSCT in order to determine whether the procedure influenced levels of CF. Furthermore, we determined whether PASAT scoring method influenced its sensitivity in detecting CF in this particular sample. It was hypothesized that CF would be evident both pre-and post IAHSCT as evidenced by lower performance on the 2nd half of the PASAT when compared to the 1st half of the PASAT at all time points. Furthermore, given the aggressive nature of the procedure, we hypothesized that the magnitude of CF (i.e. the degree of change between 1st and 2nd half performance) would be greater for those time points immediately following the IA-HSCT procedure when compared to the magnitude of CF at baseline; with the degree of change returning to baseline levels with temporal distance from the procedure.
133
2. Methods 2.1. Participants This study was approved by the Ottawa Hospital Research Ethics Board and informed consent was obtained. A total of thirty-four individuals were screened as potential candidates for the study. Eleven individuals did not meet specific inclusion criteria; as such, twentythree individuals (14 females; 9 males) with rapidly progressing MS who failed to respond to routine therapy were enrolled. High risk of progression was defined as ≥5 relapses in the first two years of the disease or attainment of a Functional System (FS) Score of at least 3 affecting pyramidal/cerebellar subscores within 5 years of disease onset. If a patient had previously received a cytotoxic agent (Mitoxantrone or Cyclophosphamide) they must have had normal bone marrow morphology and cytogenetics before being considered eligible. Age ranged from 23 to 44 years (mean = 32.65 (5.82) years) and education ranged from completion of high school to completion of graduate degrees. Expanded Disability Status Scale (EDSS) scores pre-IA-HSCT ranged from 1.5 to 6.5 (mean = 4.87 (1.40)). Twelve individuals were diagnosed with relapsing–remitting MS (RRMS) with the remaining 11 diagnosed with secondary progressive MS (SPMS). Those with primary progressive MS were excluded. 2.2. Procedure The study was a tri-center phase II efficacy study of the role of IAHSCT on the natural history of MS. Participants underwent stem cell mobilization with IV cyclophosphamide (4.5 g/m2) and 10 days of granulocyte colony-stimulating factor (10 μg/kg/day) followed by stem cell collection using peripheral vein leukapheresis. All stem cell grafts were CD34 selected and cryopreserved until transplantation. Immunoablation was accomplished in the first six individuals using cyclophosphamide (200mg/kg), oral bulsufan (16 mg/kg), and IV rabbit antithymocyte globulin (5 mg/kg). A similar regimen was used for the remaining 17 individuals although these individuals received doseadjusted IV busulfan (9.6 mg/kg) in place of the oral bulsufan. Steroids were administered concurrently with chemotherapy. Patients did not receive further MS-disease modifying drugs or experimental therapy after HSCT. The PASAT was administered by a trained research assistant. Participants completed the 3″ and 2″ PASAT at baseline (pre-IA-HSCT) and serially every 6 months post-IA-HSCT for a period of 3 years (i.e. 36 months). PASAT performance was compared between the first and second half of the task at all time points for both the 3″ and 2″ ISIs. Responses were recorded and the following scores were obtained: total number of correct responses, total correct dyad score, and percent dyad score. According to convention, in order to have an equal number of dyad responses in each half (i.e. 30 possible dyads in both halves), a dyad score was given for a correct response to the first pair of numbers presented; despite there being no possible preceding response. Percent dyad scores were obtained using the following formula: (1− ((total correct score-dyad score)/total correct score)) × 100% [6,10]. Two participants did not complete follow-up evaluations past 18 months as they did not comply with the study demands and inclusion of their data would have made it more difficult to interpret findings in the context of the treatment of interest. 2.3. Data analysis In order to examine performance differences between the first and second half of the PASAT, data was analysed for both the 3″ and 2″ ISIs using paired-samples t-tests. This was repeated for all three scoring methods and across all time points. In order to determine the magnitude of CF at each time, difference scores were calculated such that scores from the first half of the task were subtracted from scores
134
J.A. Berard et al. / Journal of the Neurological Sciences 336 (2014) 132–137
at the second half in order to create a half-change score. In order to examine whether the magnitude of CF differed from baseline following the IA-HSCT procedure, paired-samples t-tests were used which compared half-change scores at baseline with half-change scores at all times post-IA-HSCT. This was repeated for all three scoring methods and for both the 3″ and 2″ ISIs. It should be highlighted that given the small sample size and the limited power, more sophisticated statistical analyses were not possible.
3. Results 3.1. Cognitive fatigue and scoring method sensitivity Evidence of cognitive fatigue was noted across all three scoring methods on both the 3″ and 2″ ISIs. This was reflected by a significant breakdown in task performance on the second half of the task when compared to the first half of the task (Figs. 1 & 2). Cognitive fatigue was noted at the majority of time points both pre-and post-IA-HSCT with the exception of the 3″ ISI at the 6 month follow-up and the 2″
Fig. 2. Mean 2″ PASAT scores across scoring method.
ISI at the 30 month follow-up when the total number of correct responses was examined (Table 1). Given that cognitive fatigue was noted across all three scoring methods, the way in which the PASAT is scored does not seem to influence its sensitivity in detecting CF in this particular sample. Overall, however, the dyad and percent dyad scores were more consistent in their sensitivity to CF over time as CF was noted at all time points using these two scoring methods. 3.2. Magnitude of cognitive fatigue over time Paired samples t-test analyses (Table 2) showed no significant differences in the degree of cognitive fatigue from baseline to any follow-up session post-IA-HSCT. That is, the extent to which performance declined between the first and second half of the task was comparable both before and after the procedure. These results were consistent for both the 3″ and 2″ ISIs and across all three scoring methods. 4. Discussion
Fig. 1. Mean 3″ PASAT scores across scoring method.
Consistent with previous research examining the PASAT as a measure of cognitive fatigue [4,6,10], evidence of CF was noted in our sample.
J.A. Berard et al. / Journal of the Neurological Sciences 336 (2014) 132–137
135
Table 1 Mean first and second half total correct, correct dyad, and percent dyad scores at each administration. Mean total correct (SD)
p
Mean correct dyad (SD)
p
Mean percent dyad (SD)
p
Baseline 3″ ISI – First Half 3″ ISI – Second Half 2″ ISI – First Half 2″ ISI – Second Half
21.52 (5.40) 19.78 (6.19) 17.27 (5.80) 15.14 (4.99)
.044
16.74 (7.45) 14.09 (7.92) 11.91 (6.39) 8.36 (6.09)
.023
73.25 (20.86) 64.61 (24.00) 63.65 (18.44) 48.90 (23.67)
.020
6 Month 3″ ISI – First Half 3″ ISI – Second Half 2″ ISI – First Half 2″ ISI – Second Half
23.00 (6.13) 22.50 (5.52) 17.14 (6.42) 15.27 (5.82)
.460
19.23 (8.18) 16.82 (7.80) 11.77 (8.06) 8.64 (6.88)
.045
78.10 (22.09) 69.82 (22.45) 60.60 (24.76) 48.72 (25.26)
.007
12 Month 3″ ISI – First Half 3″ ISI – Second Half 2″ ISI – First Half 2″ ISI – Second Half
23.50 (6.15) 21.14 (7.04) 18.59 (5.99) 14.82 (6.28)
.006
20.00 (7.84) 16.00 (9.73) 13.00 (7.26) 8.73 (6.96)
.006
80.51 (19.92) 68.53 (24.05) 64.24 (19.79) 50.08 (24.83)
.008
18 Month 3″ ISI – First Half 3″ ISI – Second Half 2″ ISI – First Half 2″ ISI – Second Half
24.00 (5.77) 21.32 (6.62) 18.62 (6.55) 17.00 (5.44)
.001
20.68 (7.69) 16.32 (8.99) 14.05 (8.03) 10.95 (6.83)
.000
82.47 (17.25) 68.76 (26.77) 69.24 (22.02) 59.45 (21.95)
.001
24 Month 3″ ISI – First Half 3″ ISI – Second Half 2″ ISI – First Half 2″ ISI – Second Half
24.33 (5.37) 22.83 (6.30) 18.67 (7.10) 16.17 (7.10)
.019
21.28 (7.41) 18.50 (8.55) 14.33 (8.60) 10.50 (8.20)
.004
84.68 (13.69) 75.19 (22.65) 70.14 (20.98) 53.83 (28.63)
.006
30 Month 3″ ISI – First Half 3″ ISI – Second Half 2″ ISI – First Half 2″ ISI – Second Half
25.79 (4.44) 21.57 (7.61) 17.79 (5.63) 15.57 (6.81)
.019
36 Month 3″ ISI – First Half 3″ ISI – Second Half 2″ ISI – First Half 2″ ISI – Second Half
26.19 (4.89) 23.38 (7.01) 20.20 (4.72) 17.00 (5.52)
.019
.035
.036
.000
.038
.000
.102
.003
.003
.002
.000
.001
.000
23.14 (6.51) 17.50 (9.83) 12.57 (6.72) 9.36 (7.58)
.005
23.88 (7.34) 19.19 (10.80) 15.27 (6.62) 10.73 (6.79)
.014
These results support the notion that the PASAT is a reliable and sensitive measure of CF in MS [10,11]. A significant breakdown in task performance was noted on the second half of the task when compared to the first half; indicative of an inability of our sample to maintain the required cognitive effort necessary to continuously and successfully meet the task demands over time (Figs. 1 & 2). These results were consistent at both the 3″ and 2″ ISIs; thus CF was apparent regardless of overall task complexity. Previous work which examined the relationship between task complexity and levels of CF in a sample of early-phase RRMS found that, in contrast to results from the current study, CF was not apparent on the 2″ ISI [6]. It was suggested that the lack of CF noted at this more difficult ISI may have been due to overall levels of poor performance across the task (i.e. there was no breakdown in performance noted over time as performance was already quite low to begin with). In contrast, the current sample showed a trend for greater initial performance on the first half of the task when compared to the previous study and as such, a measureable breakdown in performance (i.e. CF) occurred between the first and second half of the task. The inconsistency between these two results may be due to differences in the disease characteristics of the individuals comprising the two samples given that our previous study focused on a homogenous sample of early-phase RRMS, whereas the current study focused on individuals with rapidly progressive RRMS or SPMS with poor prognosis. One can speculate that the vulnerability to CF noted on the 2″ ISI in the current sample may be attributed to the differences in disease subtypes or disease durations as compared to our previous study. To date, however, little research has examined differences in susceptibility to CF between individuals with differing subtypes of MS, as well as between those with longer versus shorter disease durations.
.033
.002
.000
.003
.000
.004
.000
87.53 (14.20) 73.74 (24.78) 65.14 (20.58) 50.04 (27.77)
.006
87.77 (20.12) 73.49 (31.82) 72.65 (17.29) 57.09 (23.56)
.021
.006
.001
Given the small sample size in the current study, it was not possible to examine differences in disease characteristics in this particular sample. Our conclusions, therefore, are purely speculation. As such, future studies should examine the impact of disease subtype (as well as in those who present with clinically isolated syndrome) and disease duration on levels of CF in order to determine whether scoring methodology impacts PASAT sensitivity to CF between these groups. Previous research has shown that the percent dyad scoring method was the most sensitive measure of CF on the PASAT [6]. Results from the current study, however, showed no greater sensitivity with one scoring method over the others. Although the dyad and percent dyad scores yielded more consistent findings across time, cognitive fatigue was noted across all three scoring methods; suggesting that, unlike previous work, in this sample a measure of performance strategy (i.e. percent dyad scores) does not appear to be more sensitive to CF when compared to measures of performance level (i.e. total number correct and total correct dyads). We can speculate that this inconsistency between the two studies may again be attributed to differences in disease characteristics between the two samples as discussed above. Cognitive fatigue was noted at all time points both pre- and post-IAHSCT, with the exception of the 3″ ISI at the 6month follow-up and the 2″ ISI at the 30 month follow-up when the total number of correct responses was examined (Table 1). While these two anomalous findings suggest that performance remained relatively stable throughout the task at these times (i.e. there was a lack of CF), this anomaly is unlikely to be clinically meaningful given the overall pattern for these individuals to exhibit cognitive fatigue at all other times, as well as the findings supporting CF at these time points when using the other scoring methods. Generally, individuals in this sample were vulnerable to CF
136
J.A. Berard et al. / Journal of the Neurological Sciences 336 (2014) 132–137
Table 2 Degree of change in cognitive fatigue from baseline to all follow-up sessions post-IA-HSCT. Mean total correct (SD)
p
Mean correct dyad (SD)
p
Mean percent dyad (SD)
p
Baseline to 6 Months 3″ ISI – Half-Change @ Baseline 3″ ISI – Half-Change @ 6 Months 2″ ISI – Half-Change @ Baseline 2″ ISI – Half-Change @ 6 Months
1.77 (3.99) .50 (3.11) 2.14 (4.43) 1.86 (3.91)
.172
2.77 (5.30) 2.41 (5.29) 3.55 (4.88) 3.14 (4.27)
.787
9.09 (16.70) 8.28 (13.03) 14.75 (15.52) 11.88 (16.36)
.839
Baseline to 12 Months 3″ ISI – Half-Change @ Baseline 3″ ISI – Half-Change @ 12 Months 2″ ISI – Half-Change @ Baseline 2″ ISI – Half-Change @ 12 Months
1.68 (3.98) 2.36 (3.66) 1.95 (4.46) 3.95 (3.31)
.538
2.50 (5.28) 4.00 (6.07) 3.38 (4.93) 4.48 (4.02)
.387
Baseline to 18 Months 3″ ISI – Half-Change @ Baseline 3″ ISI – Half-Change @ 18 Months 2″ ISI – Half-Change @ Baseline 2″ ISI – Half-Change @ 18 Months
1.55 (3.88) 2.68 (3.40) 2.10 (4.54) 1.62 (3.34)
.341
2.27 (5.00) 4.36 (4.77) 3.62 (4.98) 3.10 (3.77)
.174
Baseline to 24 Months 3″ ISI – Half-Change @ Baseline 3″ ISI – Half-Change @ 24 Months 2″ ISI – Half-Change @ Baseline 2″ ISI – Half-Change @ 24 Months
.72 (3.43) 1.50 (2.46) 1.71 (4.62) 2.41 (2.29)
.454
1.33 (4.92) 2.78 (3.49) 3.18 (5.21) 3.94 (3.70)
.307
Baseline to 30 Months 3″ ISI – Half-Change @ Baseline 3″ ISI – Half-Change @ 30 Months 2″ ISI – Half-Change @ Baseline 2″ ISI – Half-Change @ 30 Months
1.07 (3.43) 4.21 (5.86) 2.31 (5.34) 2.00 (4.83)
.095
1.79 (5.04) 5.64 (6.25) 3.77 (5.91) 3.31 (5.25)
.120
Baseline to 36 Months 3″ ISI – Half-Change @ Baseline 3″ ISI – Half-Change @ 36 Months 2″ ISI – Half-Change @ Baseline 2″ ISI – Half-Change @ 36 Months
1.38 (3.42) 2.81 (4.26) 2.20 (4.96) 3.20 (3.41)
.845
.104
.701
.619
.890
.292 .531
both before and after the IA-HSCT procedure. It is of note, however, that the procedure itself did not lead to any greater vulnerability to CF given that individuals showed a similar magnitude of CF both pre-and post-IAHSCT (Table 2). That is, the degree in which task performance declined across the task was comparable at all time points, suggesting the procedure itself (or presumably the associated atrophy) does not seem to negatively impact an individual’s susceptibility to CF. As seen in Figs. 1 & 2, there was a trend for subjects to perform better on the PASAT with each successive administration; consistent with the practice effects that are usually seen with repeat testing [10,17,18]. As part of the larger overall project which includes the current study, we examined the impact of the IA-HSCT procedure on overall PASAT performance and found that after taking practice effects into account (using reliable change methodology) there still remained an overall trend for improvement in performance (while non-significant) across the follow-up sessions [19]. It is notable that despite this overall trend for increased performance, individuals' vulnerability to CF did not show a similar pattern of improvement over time. Individuals in this sample continued to remain vulnerable to CF across the follow-up sessions when performing a sustained and cognitively demanding task (as evidenced by a breakdown in their performance across the task) despite their overall increase in performance levels. At this time, it is difficult to say if the lack of impact of the IA-HSCT procedure on levels of CF would also be noted using traditional treatment therapies. While studies have examined the relationship between some traditional therapies and levels of overall fatigue in MS [20,21], few studies have examined the relationship between these treatments and CF specifically. As such, future studies should examine changes in CF following these traditional therapies in order to determine how they compare with those changes seen following the IA-HSCT procedure. Along with previous work by our group examining the impact of the IA-HSCT procedure on measures of cognition and neuroimaging [16,19],
2.25 (5.03) 4.69 (6.71) 3.60 (5.50) 4.53 (4.47)
.775
.391
.737
.659
.843
.207 .625
8.31 (16.76) 11.98 (19.22) 14.40 (15.82) 14.84 (13.73)
6.64 (13.67) 13.72 (15.92) 15.43 (15.58) 9.79 (13.78)
3.58 (12.51) 9.49 (12.81) 13.55 (14.95) 14.91 (11.72)
4.48 (13.79) 13.79 (15.72) 15.21 (16.43) 14.34 (17.79)
5.22 (13.18) 14.27 (22.18) 14.23 (15.44) 15.57 (14.16)
.515
.472 .921
.136 .304
.160 .767
.160 .896
.122 .799
the current study highlights the promise of the IA-HSCT procedure as a viable treatment option for individuals with MS who do not respond well to traditional treatment therapies given it seems the procedure itself does not further exacerbate CF. While results suggest that the procedure itself does not ameliorate an individual's susceptibility to CF; neither does it seem to negatively impact levels of CF despite its aggressive nature. One limitation of the current study however is the lack of a control group. While attempts were made to recruit a suitable control group comprising individuals undergoing bone marrow transplant for other indications besides MS, attempts were unsuccessful. Although we conclude that the IA-HSCT procedure had no impact on individuals' level of CF and that individuals remained vulnerable to CF despite a trend for overall improvement in performance, a control group would have allowed us to examine whether these results were specific to individuals with rapidly progressive RRMS or SPMS undergoing IA-HSCT. Overall, while there seems to be no detrimental impact on levels of CF for those individuals who have undergone IA-HSCT, CF does continue to present as an ongoing problem for these individuals post-procedure. Evidence from the current study suggests that, despite the apparent improvements in PASAT performance over time, individuals continue to show a breakdown in task performance indicative of CF as the task progresses. These findings have important clinical relevance for the study of CF when one considers that individuals may present with apparently preserved cognitive performance, but may subjectively report feeling as though they have to mentally work harder in order to achieve or maintain this level of performance. Functional neuroimaging studies would allow for the evaluation of changes in neural activity across the task, and may provide objective evidence to substantiate these subjective reports. While structural neuroimaging variables were collected for the current participants, functional neuroimaging was not performed. The fact that cognitive performance seems to improve (albeit to a non-significant degree) following IA-HSCT, but cognitive
J.A. Berard et al. / Journal of the Neurological Sciences 336 (2014) 132–137
fatigue does not, suggests that these two concepts are distinct, and provides confirmatory evidence for anecdotal reports from individuals with MS who say they can often maintain adequate levels of cognitive performance compared to their peers, but only at the cost of high levels of cognitive fatigue and mental effort. In conclusion, despite the fact that no improvements in cognitive fatigue were noted over time, the current findings provide important safety information about IA-HSCT, in that the procedure appears to have no detrimental impact on levels of cognitive fatigue, either immediately post-procedure, or up to three years post-procedure. This information, coupled with the findings of a potential mild benefit on cognitive performance, lends further credence to the use of IA-HSCT given that the procedure, despite its aggressive nature, does not seem to further exacerbate CF. Conflict of interest statement The authors declare that they have no conflict of interest. Acknowledgments We would like to gratefully thank the participants of this study for their time and effort. This research was supported by a grant from the Research Foundation of the Multiple Sclerosis Society of Canada. References [1] Kinkel RP. Fatigue in multiple sclerosis: reducing the impact through comprehensive management. Int J MS Care 2000;2(3):3–10. [2] Amato MP, Ponziani G, Rossi F, Liedl CL, Stefanile C, Rossi L. Quality of life in multiple sclerosis: the impact of depression, fatigue, and disability. Mult Scler 2001;7:340–4. [3] Schwid SR, Covington M, Segal BM, Goodman AD. Fatigue in multiple sclerosis: current understanding and future directions. J Rehabil Res Dev 2002;39:211–24. [4] Bryant M, Chiaravalloti ND, DeLuca J. Objective measurement of cognitive fatigue in multiple sclerosis. Rehabil Psychol 2004;49:114–22. [5] Tombaugh TN. A comprehensive review of the Paced Auditory Serial Addition Test. Arch Clin Neuropsychol 2006;21:53–76.
137
[6] Walker LAS, Berard JA, Berrigan LI, Rees LM, Freedman MS. Detecting cognitive fatigue in multiple sclerosis: method matters. J Neurol Sci 2012;316:86–92. [7] Walker LAS, Cheng A, Berard JA, Berrigan LI, Rees LM, Freedman MS. Tests of information processing speed – what do people with multiple sclerosis think about them? Int J MS Care 2012;14:92–9. [8] Sl Barker-Collo. Within session practice effects on the PASAT in clients with multiple sclerosis. Arch Clin Neuropsychol 2005;20:145–52. [9] Brittain JL, LaMarche JA, Reeder KP, Roth DL, Boll TJ. Effects of age and IQ on Paced Auditory Serial Addition Test (PASAT) performance. Clin Neuropsychol 1991;5:163–75. [10] Fisk JD, Archibald CJ. Limitations of the Paced Auditory Serial Addition Test (PASAT) as a measure of working memory in patients with multiple sclerosis. J Int Neuropsychol Soc 2001;7:363–72. [11] Forn C, Belenguer A, Parcet-Ibars MA, Avila C. Information-processing speed is the primary deficit underlying the poor performance of multiple sclerosis patients in the Paced Auditory Serial Addition Test (PASAT). J Clin Exp Neuropsychol 2008;30:789–96. [12] Schwid SR, Tyler CM, Scheid EA, Weinstein A, Goodman AD, McDermott MP. Cognitive fatigue during a test requiring sustained attention: a pilot study. Mult Scler 2003;9:503–8. [13] Pasquini MC, Griffith LM, Arnold DL, Atkins HL, Bowen JD, Chen JT, et al. Hematopoietic stem cell transplantation for multiple sclerosis: collaboration of the CIBMTR and EBMT to facilitate international clinical studies. Biol Blood Marrow Transplant 2010;16(8):1076–83. [14] Freedman MS. Bone marrow transplantation: does it stop MS progression? J Neurol Sci 2007;259:85–9. [15] Chen JT, Collins DL, Atkins HL, Freedman MS, Galai A, Arnold DL, et al. Brain atrophy after immunoablation and stem cell transplantation in multiple sclerosis. Neurology 2006;66:1935–7. [16] Walker LAS, Berard JA, Bowman M, Atkins HL, Lee H, Freedman MS. Cognitive change and neuroimaging following immunoablative therapy and hematopoietic stem cell transplantation in multiple sclerosis: a pilot study. Multi Scler Relat Disord 2013. http://dx.doi.org/10.1016/j.msard.2013.05.001. [17] Solari A, Radice D, Manneschi L, Motti L, Montanari E. The multiple sclerosis functional composite: different practice effects in the three test components. J Neurol Sci 2005;228:71–4. [18] McCaffrey RJ, Ortega A, Orsillo SM, Nelles WB, Haase RF. Practice effects in repeated neuropsychological assessments. Clin Neuropsychol 1992;6:32–42. [19] Walker LAS, Berard JA, AtkinsHL, Bowman M, LeeH, FreedmanMS. Longitudinal change in PASAT performance following immunoablative therapy and haematopoietic stem cell transplant in multiple sclerosis. [Manuscript in progress]. [20] Walther EV, Hohfeld R. Multiple sclerosis: side effects of interferon beta therapy and their management. Neurology 1999;10:1622–7. [21] Putzki N, Yaldizli Ö, Tettenborn B, Diener HC. Multiple sclerosis associated fatigue during natalizumab treatment. J Neurol Sci 2009;285:109–13.
