Episodic Memory Impairments in Primary Brain Tumor Patients

Episodic Memory Impairments in Primary Brain Tumor Patients Abstract Objective Cognitive investigations in brain tumor patients have mostly explored episodic memory without differentiating between encoding, storage, and retrieval deficits. The aim of this study is to offer insight into the memory sub-processes affected in primary brain tumor patients and propose an appropriate assessment method. Method We retrospectively reviewed the clinical and memory assessments of 158 patients with primary brain tumors who had presented to our departments with cognitive complaints and were investigated using the Free and Cued Selective Reminding Test. Results Retrieval was the process of episodic memory most frequently affected, with deficits in this domain detected in 92% of patients with episodic memory impairments. Storage and encoding deficits were less prevalent, with impairments, respectively, detected in 41% and 23% of memory-impaired patients. The pattern of episodic memory impairment was similar across different tumor histologies and treatment modalities. Conclusion Although all processes of episodic memory were found to be impaired, retrieval was by far the most widely affected function. A thorough assessment of all three components of episodic memory should be part of the regular neuropsychological evaluation in patients with primary brain tumors. Memory decay, Brain neoplasms, Neurotoxicity, Cognitive assessment Introduction Cognitive impairments are a frequent and disabling complication in patients treated for brain tumors (Klein, 2015). Cognitive deficits can be related to tumor infiltration, tumor-related epilepsy, or to the effects of antitumoral treatments such as chemo- and radiotherapy. Medications such as antiepileptic drugs and steroids, which are frequently administered to these patients, have also been suggested to negatively influence cognitive function (Klein, 2015). Despite their dramatic impact on patients’ quality of life (Giovagnoli et al., 2014), the reported incidence of cognitive deficits in primary brain tumor patients varies widely across studies, from 20% to more than 80% (van Loon et al., 2015). Tests used and delays between diagnosis and assessment may partially explain these variations (Douw et al., 2009). Memory and attention have been consistently reported as the cognitive functions most frequently affected in brain tumor patients (Klein, 2015). Memory is a highly complex cognitive process that is normally split between different sub-process according to the nature of information being processed (e.g., episodic, semantic, procedural). Noll and colleagues recently showed that tumor location can specifically affect subtle cognitive processes and lead to distinct patterns of episodic memory dysfunction (Noll, Weinberg, Ziu, & Wefel, 2016). Episodic memory can be defined as the memory of personally experienced events within their temporo-spatial context: it allows for conscious recall of past experience of an event (what), its location (where), the time it occurred (when), and its emotional content (Tulving, 2004). Hermann Ebbinghaus (1850–1909) first introduced the concepts of encoding, storing/consolidation, and retrieval. These three sub-processes are the key phases in long-term memory formation and are essential for retrieving information when required. Encoding is the first phase of long-term memorization and consists of processing the characteristics of the information to be memorized in association with their environmental, cognitive, and emotional context. Storage refers to the consolidation of the encoded information in the memory system in order to prevent long-term forgetting. Retrieval is the final active phase in which the cognitive system transfers the information from long-term memory to working memory. The three sub-processes of episodic memory can take place alone or in combination, as has been highlighted in various pathological conditions (Lemos, Duro, Simões, & Santana, 2014). Therefore, an exhaustive investigation of episodic memory should include an assessment of all three of its components in order to detect which sub-processes are affected. So far, investigations in neuro-oncology have primarily explored episodic memory performance, but few studies have taken an interest in the mechanisms involved. The aim of this study is to offer insight into episodic memory sub-processes affected in primary brain tumor patients. Methods We performed a retrospective search of our institutional databases for patients with primary brain tumors referred to our departments (HIA de Percy, Clamart; GH Pitié-Salpêtrière, Paris, France) for suspected tumor- or treatment-related cognitive impairment during the period 2010–2015. Inclusion criteria were (1) a history of primary brain tumor with one of the following histologies: low-grade glioma (LGG), high-grade glioma (HGG), meningioma, or primary central nervous system lymphoma (PCNSL); (2) age ≥ 18 years old; and (3) episodic memory assessment performed using the French version of the Free and Cued Selective Reminding Test (van der Linden et al., 2004). Exclusion criteria were (1) the coexistence of other neurological or psychiatric disorders affecting cognitive function; and (2) personal history of a second neoplasm. Out of a total of 192 patients evaluated in our departments during this period, 158 matched the study criteria and could be included in this report. Clinical and neuropsychological assessments of the patients included in the study were collected and reviewed in detail. This research was carried out in accordance with the principles of the Declaration of Helsinki. Episodic Memory Assessment Verbal episodic memory was assessed using the Free and Cued Selective Reminding Test (FCSRT) (van der Linden et al. 2004). The FCSRT is widely used in francophone countries and has demonstrated its ability to detect and discriminate between encoding, storage, and retrieval impairments (Lemos et al., 2014). The test was adapted from the paradigm proposed by Buschke (Buschke, 1984) and Grober (Grober & Buschke, 1987) and is designed to manipulate encoding and retrieval conditions through the use of cues. It was developed on the basis of the observation that cues improve retrieval, especially when used not only during retrieval, but also during encoding (Tulving & Thomson, 1973). The FCSRT consists of several tasks. The first is a controlled encoding task during which the patient is asked to learn 16 words and to associate each with its semantic category (e.g., the word “judo” is associated with the category “sport”). It is followed by a free and cued recall task. During free recall, the patient is asked to give as many words as possible from the list of 16 words. During cued recall, the patient is provided with the category given at the encoding phase for each item not retrieved freely (e.g., what was the sport?). This is followed by a recognition task during which the patient is asked to identify the 16 original words, and reject distractors, from a list of 48 words. The test concludes with a delayed free and cued recall task 20 min after the end of the recognition task. The test is scored based on the number of items recalled freely compared to the total number of items retrieved (free and cued recall). Between each recall task, the patient must perform a distraction task, consisting of counting for 20 s. Patients’ scores in our study were compared with normative data from 483 healthy controls (Van der Linder et al., 2004) matched for age, gender, and education. Each of the nine scores recorded (number of word recalled at immediate recall), free recalls (1, 2, 3, delayed) and cued recalls (1, 2, 3, delayed) were considered abnormal when it corresponded to a performance equal to or under the fifth percentile of the healthy controls normative data (van der Linden et al., 2004). An encoding deficit was diagnosed when the immediate recall was abnormal (the assumption being that the items were not present in the working memory immediately after they have been red, and so not encoded). A failure in free recalls corresponded to at least 2/4 abnormal scores and a failure in cued recalls corresponded to at least 2/4 abnormal scores (the test is composed of four free and four cued recalls). A storage deficit was diagnosed in the case of a failure in free recalls associated with a failure in cued recalls. This means that the cues didn’t improve the number of items recalled, assuming that the items was not stored. A retrieval deficit was diagnosed in the case of a failure in free recalls isolated (normal cued recalls). Indeed, the cues improved the number of items recalled comparably to healthy controls, giving a proof that items was stored in the memory but not available at the moment. Furthermore, an association of storage and retrieval deficit was diagnosed in the case of a failure in free recalls and a failure in cued recalls, but with limited improvement (incomparably to healthy controls) of the total number of items recalled with cues. In other words, items were retrieved with difficulties (because cues helped after a failure in free recalls), but was also not stored as expected (because cues did not help to reach normal number of recalls, meaning that items was not entirely stored). To help in the understanding of the FCRST design, Supplementary Fig. 1 gives an overview of the protocol, the criteria and the interpretation of the different pattern of results. Statistical Analysis Descriptive analyses were used for FCSRT scores and for patient and tumor characteristics. Patients with episodic memory impairments were compared to non-impaired patients to search for differences in demographic or tumor characteristics. The t-test was used for continuous variables. The χ2 or Fisher’s exact test was used for categorical comparison. Univariate and multiple logistic regressions were used to explore the association between the presence of episodic memory deficits and a number of patient and tumor characteristics. All data analyses were performed using R statistical software. Results Clinical characteristics of our 158 patients are summarized in Table 1. Median age at the time of neuropsychological assessment was 58 (range 22–87). Tumor histology was as follows: HGG in 79 patients (50%), LGG in 30 patients (19%), meningioma in 33 patients (21%), and PCNSL in 16 patients (10%). All patients had supratentorial neoplasms, the frontal lobes being the most frequent location. Tumor treatment included surgery, radiotherapy, and/or chemotherapy using various schemes and combinations, based on tumor histology and the patient’s clinical conditions. Overall, 78 patients received surgical resection (50%), 64 patients were irradiated (40%), and another 64 received chemotherapy (40%). Table 1. Patients’ and tumor characteristics Demographics  Whole group (n = 158)  Non-impaired group (n = 92)  Impaired group (n = 66)  p-valuea  Male/female  86 (54%)/72 (46%)  50 (54%)/42 (46%)  36 (54%)/30 (46%)  1  Age (median, range)  58 (22–87)  55 (26–87)  62 (22–87)  .082  Education   <12 years  40 (25%)  26 (28%)  14 (21%)  .556   12 years  20 (13%)  12 (13%)  8 (12%)     >12 years  98 (62%)  54 (59%)  44 (67%)    Handedness   Left  5 (3%)  2 (2%)  3 (5%)  .045   Ambidextrous  6 (4%)  6 (6%)  0     Right  147 (93%)  84 (92%)  63 (95%)    Tumor characteristics   Histology    LGG  30 (19%)  9 (10%)  21 (32%)  .001    HGG  79 (50%)  48 (52%)  31 (47%)      Meningioma  33 (21%)  27 (29%)  6 (9%)      PCNSL  16 (10%)  8 (9%)  8 (12%)    Locationb   Left frontal  43 (27%)  21 (23%)  22 (33%)  .04   Right frontal  99 (62%)  70 (76%)  29 (44%)     Left temporal  18 (11%)  9 (10%)  9 (14%)     Right temporal  35 (22%)  25 (27%)  10 (15%)     Left parietal  12 (8%)  6 (6%)  6 (9%)     Right parietal  16 (10%)  8 (9%)  8 (12%)     Left occipital  4 (2%)  3 (3%)  1 (1%)     Right occipital  13 (8%)  11 (12%)  2 (2%)     Corpus Callosum  15 (9%)  6 (6%)  9 (14%)     Basal ganglia  16 (10%)  8 (9%)  8 (12%)    Prior antitumoral treatmentsc   Biopsy alone  40 (25%)  21 (23%)  19 (29%)  .013   Exeresis alone  41 (26%)  33 (36%)  8 (12%)     RT alone  6 (4%)  4 (4%)  2 (3%)     CT alone  11 (7%)  8 (9%)  3 (4%)     Exeresis + CT  2 (1%)  0  2 (3%)     Exeresis + RT  7 (4%)  4 (4%)  3 (4%)     RT + CT  23 (14%)  10 (11%)  13 (20%)     Exeresis + RT + CT  28 (18%)  12 (13%)  16 (24%)     Delay since RT (mean, SD)  7.9 years (±6.6)  7.7 years (±6.2)  8.2 years (±8.3)  .835   Delay since RT (range)  0.5–20 years  0.5–12 years  0.5–20 years    Concomitant medications   Antiepileptic drugs  91 (58%)  51 (55%)  40 (61%)  .627   Steroids  35 (22%)  22 (24%)  13 (20%)  .663  Demographics  Whole group (n = 158)  Non-impaired group (n = 92)  Impaired group (n = 66)  p-valuea  Male/female  86 (54%)/72 (46%)  50 (54%)/42 (46%)  36 (54%)/30 (46%)  1  Age (median, range)  58 (22–87)  55 (26–87)  62 (22–87)  .082  Education   <12 years  40 (25%)  26 (28%)  14 (21%)  .556   12 years  20 (13%)  12 (13%)  8 (12%)     >12 years  98 (62%)  54 (59%)  44 (67%)    Handedness   Left  5 (3%)  2 (2%)  3 (5%)  .045   Ambidextrous  6 (4%)  6 (6%)  0     Right  147 (93%)  84 (92%)  63 (95%)    Tumor characteristics   Histology    LGG  30 (19%)  9 (10%)  21 (32%)  .001    HGG  79 (50%)  48 (52%)  31 (47%)      Meningioma  33 (21%)  27 (29%)  6 (9%)      PCNSL  16 (10%)  8 (9%)  8 (12%)    Locationb   Left frontal  43 (27%)  21 (23%)  22 (33%)  .04   Right frontal  99 (62%)  70 (76%)  29 (44%)     Left temporal  18 (11%)  9 (10%)  9 (14%)     Right temporal  35 (22%)  25 (27%)  10 (15%)     Left parietal  12 (8%)  6 (6%)  6 (9%)     Right parietal  16 (10%)  8 (9%)  8 (12%)     Left occipital  4 (2%)  3 (3%)  1 (1%)     Right occipital  13 (8%)  11 (12%)  2 (2%)     Corpus Callosum  15 (9%)  6 (6%)  9 (14%)     Basal ganglia  16 (10%)  8 (9%)  8 (12%)    Prior antitumoral treatmentsc   Biopsy alone  40 (25%)  21 (23%)  19 (29%)  .013   Exeresis alone  41 (26%)  33 (36%)  8 (12%)     RT alone  6 (4%)  4 (4%)  2 (3%)     CT alone  11 (7%)  8 (9%)  3 (4%)     Exeresis + CT  2 (1%)  0  2 (3%)     Exeresis + RT  7 (4%)  4 (4%)  3 (4%)     RT + CT  23 (14%)  10 (11%)  13 (20%)     Exeresis + RT + CT  28 (18%)  12 (13%)  16 (24%)     Delay since RT (mean, SD)  7.9 years (±6.6)  7.7 years (±6.2)  8.2 years (±8.3)  .835   Delay since RT (range)  0.5–20 years  0.5–12 years  0.5–20 years    Concomitant medications   Antiepileptic drugs  91 (58%)  51 (55%)  40 (61%)  .627   Steroids  35 (22%)  22 (24%)  13 (20%)  .663  LGG = low-grade glioma; HGG = high-grade glioma; PCNSL = primary central nervous system lymphoma; RT = radiotherapy; CT = chemotherapy. at-Tests, Pearson χ2 or Fisher’s exact test for comparisons between non-impaired and impaired group. bAll patients had a unique tumor but 51 patients had a tumor involving several brain lobes. cRadiotherapy, whenever performed, was done using a 3D-conformational technique. Table 1. Patients’ and tumor characteristics Demographics  Whole group (n = 158)  Non-impaired group (n = 92)  Impaired group (n = 66)  p-valuea  Male/female  86 (54%)/72 (46%)  50 (54%)/42 (46%)  36 (54%)/30 (46%)  1  Age (median, range)  58 (22–87)  55 (26–87)  62 (22–87)  .082  Education   <12 years  40 (25%)  26 (28%)  14 (21%)  .556   12 years  20 (13%)  12 (13%)  8 (12%)     >12 years  98 (62%)  54 (59%)  44 (67%)    Handedness   Left  5 (3%)  2 (2%)  3 (5%)  .045   Ambidextrous  6 (4%)  6 (6%)  0     Right  147 (93%)  84 (92%)  63 (95%)    Tumor characteristics   Histology    LGG  30 (19%)  9 (10%)  21 (32%)  .001    HGG  79 (50%)  48 (52%)  31 (47%)      Meningioma  33 (21%)  27 (29%)  6 (9%)      PCNSL  16 (10%)  8 (9%)  8 (12%)    Locationb   Left frontal  43 (27%)  21 (23%)  22 (33%)  .04   Right frontal  99 (62%)  70 (76%)  29 (44%)     Left temporal  18 (11%)  9 (10%)  9 (14%)     Right temporal  35 (22%)  25 (27%)  10 (15%)     Left parietal  12 (8%)  6 (6%)  6 (9%)     Right parietal  16 (10%)  8 (9%)  8 (12%)     Left occipital  4 (2%)  3 (3%)  1 (1%)     Right occipital  13 (8%)  11 (12%)  2 (2%)     Corpus Callosum  15 (9%)  6 (6%)  9 (14%)     Basal ganglia  16 (10%)  8 (9%)  8 (12%)    Prior antitumoral treatmentsc   Biopsy alone  40 (25%)  21 (23%)  19 (29%)  .013   Exeresis alone  41 (26%)  33 (36%)  8 (12%)     RT alone  6 (4%)  4 (4%)  2 (3%)     CT alone  11 (7%)  8 (9%)  3 (4%)     Exeresis + CT  2 (1%)  0  2 (3%)     Exeresis + RT  7 (4%)  4 (4%)  3 (4%)     RT + CT  23 (14%)  10 (11%)  13 (20%)     Exeresis + RT + CT  28 (18%)  12 (13%)  16 (24%)     Delay since RT (mean, SD)  7.9 years (±6.6)  7.7 years (±6.2)  8.2 years (±8.3)  .835   Delay since RT (range)  0.5–20 years  0.5–12 years  0.5–20 years    Concomitant medications   Antiepileptic drugs  91 (58%)  51 (55%)  40 (61%)  .627   Steroids  35 (22%)  22 (24%)  13 (20%)  .663  Demographics  Whole group (n = 158)  Non-impaired group (n = 92)  Impaired group (n = 66)  p-valuea  Male/female  86 (54%)/72 (46%)  50 (54%)/42 (46%)  36 (54%)/30 (46%)  1  Age (median, range)  58 (22–87)  55 (26–87)  62 (22–87)  .082  Education   <12 years  40 (25%)  26 (28%)  14 (21%)  .556   12 years  20 (13%)  12 (13%)  8 (12%)     >12 years  98 (62%)  54 (59%)  44 (67%)    Handedness   Left  5 (3%)  2 (2%)  3 (5%)  .045   Ambidextrous  6 (4%)  6 (6%)  0     Right  147 (93%)  84 (92%)  63 (95%)    Tumor characteristics   Histology    LGG  30 (19%)  9 (10%)  21 (32%)  .001    HGG  79 (50%)  48 (52%)  31 (47%)      Meningioma  33 (21%)  27 (29%)  6 (9%)      PCNSL  16 (10%)  8 (9%)  8 (12%)    Locationb   Left frontal  43 (27%)  21 (23%)  22 (33%)  .04   Right frontal  99 (62%)  70 (76%)  29 (44%)     Left temporal  18 (11%)  9 (10%)  9 (14%)     Right temporal  35 (22%)  25 (27%)  10 (15%)     Left parietal  12 (8%)  6 (6%)  6 (9%)     Right parietal  16 (10%)  8 (9%)  8 (12%)     Left occipital  4 (2%)  3 (3%)  1 (1%)     Right occipital  13 (8%)  11 (12%)  2 (2%)     Corpus Callosum  15 (9%)  6 (6%)  9 (14%)     Basal ganglia  16 (10%)  8 (9%)  8 (12%)    Prior antitumoral treatmentsc   Biopsy alone  40 (25%)  21 (23%)  19 (29%)  .013   Exeresis alone  41 (26%)  33 (36%)  8 (12%)     RT alone  6 (4%)  4 (4%)  2 (3%)     CT alone  11 (7%)  8 (9%)  3 (4%)     Exeresis + CT  2 (1%)  0  2 (3%)     Exeresis + RT  7 (4%)  4 (4%)  3 (4%)     RT + CT  23 (14%)  10 (11%)  13 (20%)     Exeresis + RT + CT  28 (18%)  12 (13%)  16 (24%)     Delay since RT (mean, SD)  7.9 years (±6.6)  7.7 years (±6.2)  8.2 years (±8.3)  .835   Delay since RT (range)  0.5–20 years  0.5–12 years  0.5–20 years    Concomitant medications   Antiepileptic drugs  91 (58%)  51 (55%)  40 (61%)  .627   Steroids  35 (22%)  22 (24%)  13 (20%)  .663  LGG = low-grade glioma; HGG = high-grade glioma; PCNSL = primary central nervous system lymphoma; RT = radiotherapy; CT = chemotherapy. at-Tests, Pearson χ2 or Fisher’s exact test for comparisons between non-impaired and impaired group. bAll patients had a unique tumor but 51 patients had a tumor involving several brain lobes. cRadiotherapy, whenever performed, was done using a 3D-conformational technique. Patients were evaluated at variable intervals after the tumor diagnosis (range 6–240 months, median 92). At the time of cognitive assessment, 58% of patients were taking antiepileptic drugs and 22% were taking steroids; none of the patients was actively receiving antitumoral treatments as their tumors were considered in remission. Overall, 66 (42%) of the patients in our study series were found to have episodic memory impairments. Retrieval deficits were by far the most frequent impairment, affecting 39% of patients in the whole cohort and 92% of patients with episodic memory impairments (isolated or in association with encoding and/or storage deficits). Encoding and storage deficits were less prevalent, respectively, affecting 9% and 17% of patients in the whole cohort, and 23% and 41% of patients with episodic memory impairments (Fig. 1A). Encoding deficits were associated with both storage (Logistic regression: OR 5.38, β = 1.68, 95% IC 2.07–13.85, p = .003) and retrieval (Logistic regression: OR 3.6, β = 1.28, 95% IC 1.44–9.82, p = .02) deficits. Fig. 1B shows the frequencies of encoding, storage, and retrieval impairments across different tumor histologies. Fig. 1. View largeDownload slide Frequency of encoding, storage, and retrieval deficits. (A) Frequency of encoding, storage, and retrieval impairments in the whole cohort (light gray) and in the impaired patient group (dark gray). (B) frequency of encoding (light gray), storage (medium gray), and retrieval (dark gray) impairments based on tumor histology (whole cohort). Fig. 1. View largeDownload slide Frequency of encoding, storage, and retrieval deficits. (A) Frequency of encoding, storage, and retrieval impairments in the whole cohort (light gray) and in the impaired patient group (dark gray). (B) frequency of encoding (light gray), storage (medium gray), and retrieval (dark gray) impairments based on tumor histology (whole cohort). We then compared patients with episodic memory impairments (impaired patients) to patients without any episodic memory impairments (non-impaired patients) (Table 1). The group of impaired patients had a higher prevalence of LGG and a lower prevalence of meningioma histology (χ2: p = .001). Impaired patients were also more likely to have received radio-chemotherapy (concomitantly ± sequentially, or only sequentially) and less likely to have undergone exeresis alone (χ2: p = .013). Logistic regression confirmed that a histology of LGG (Logistic regression: OR 4.304, β = 1.459, 95% IC = 1.86–10.63, p = .0009) and prior radio-chemotherapy treatment (Logistic regression: OR 2.494, β = 0.914, 95% IC 1.26–4.98, p = .008) were factors significantly associated with memory impairment, after adjusting for other patient and tumor characteristics. No correlations between specific tumor locations and types of episodic memory impairment were found, except for the association of encoding deficits with corpus callosum infiltration (Logistic regression: OR 4.36, β = 1.68, 95% IC 1.37–12.58, p = .02). However, when we pooled results for patients with tumors located in the left hemisphere and compared them with pooled results for patients with tumors located in the right hemisphere, we found that memory impairments appeared to be significantly more prevalent in patients with tumors located in the left hemisphere (χ2: 48% vs. 31%, p = .013). No association between the frequency or nature of episodic memory impairments and clinical characteristics such as age, gender, education, or concomitant medications was detected in our study. Discussion In the present study, we explored all three components of episodic memory in a large series of patients with primary brain tumors. Episodic memory deficits were highly prevalent in our cohort (42%), suggesting that this domain should be systematically investigated in these patients, as has already been recommended (Klein, 2015). To our knowledge, we are the first to report that retrieval was the component of episodic memory most frequently affected, with deficits in this process found in 92% of patients with episodic memory impairments (isolated or in association with encoding and/or storage deficit). Storage deficits were also present in a non-negligible proportion of patients (41% of patients with impairments). The extremely high prevalence of retrieval deficits detected in impaired patients is consistent with what we observe on a daily basis in the neuro-oncology department. Although the reasons for this finding are still unknown, we can speculate that retrieval is more frequently affected because it recruits a larger neural network (Jeong, Chung, & Kim, 2015), which is more likely to be disrupted as a result of tumor infiltration or antitumoral treatments (surgery, radiotherapy, and possibly chemotherapy). Furthermore, the frequency of impairments reported here is probably overestimated because of a recruitment bias in favor of patients presenting with cognitive complaints. These results should not be considered as thorough epidemiological data, but rather as a first overview of impairment patterns. Episodic memory can also be influenced by other cognitive processes such as selective attention, working memory, and organizational skills (Burger, Uittenhove, Lemaire, & Taconnat, 2017). One study conducted on a small group of patients with LGG addressed this particular issue, concluding that the episodic memory impairments observed in the cohort could not be attributed to deficits in these other domains (Armstrong, Stern, & Corn, 2001). Still, validation of these results in a larger prospective setting is needed. Episodic memory is a highly complex cognitive process that involves several areas and networks. It has recently been postulated that, along with the medio-temporal/hippocampal region, other areas such as the prefrontal and the parietal cortex may play an important role with regard to encoding and retrieval (Jeong et al., 2015). In our cohort, we did not observe a relation between any one single tumor location and type of episodic memory impairment, except for patients with corpus callosum infiltration who did worse at encoding, probably due to attention issues. However, patients with tumors in the left hemisphere had a significantly higher prevalence of episodic memory impairments compared to patients with tumors in the right hemisphere (χ2: p = .013). These results could be explained by the nature of the material used in the FCRST. Right structures may be more involved in spatial memory, which is not tested through the FCRST, while left structures could be more involved in context-dependent and autobiographic memory (Burgess, Maguire, & O’Keefe, 2002). Our study included patients with primary brain tumors corresponding to different histologies. We found a higher prevalence of storage and retrieval deficits in LGG patients compared to patients with other tumor types. Although tumor infiltrative behavior is an important factor to consider due to its capacity to disrupt cortical–subcortical connectivity (Duffau, 2014), the high prevalence of storage and retrieval deficits in patients with LGG observed in our series is more likely attributable to a selection bias. LGG patients live longer and, in most cases, were referred to our center for cognitive complaints after treatment (Douw et al., 2009). Still, despite these differences in prevalence, the profile of episodic memory impairments was similar across tumor types, suggesting there could be a common pathogenetic basis that transcends histology. Patients who received radio-chemotherapy appeared more prone to developing memory impairments than patients who did not. Brain irradiation remains a recognized risk factor for late neurotoxicity (Klein, 2015) due to its capacity to reduce neuronal plasticity and induce chronic progressive leukoencephalopathy (Omuro et al., 2005). Furthermore, the concomitant administration of radiation and chemotherapy could have a synergistic neurotoxic effect (Durand et al., 2015). Because of the cross-sectional design of our study, and given that most of our patients were assessed after receiving adjuvant treatments, it is impossible for us to discriminate between the effect of the tumor itself and the effect of antitumoral treatments. Further prospective studies are needed to investigate the possible influence of other cognitive functions such as attention and organizational skills (Noll et al., 2016) on memory deficits and the evolution of episodic memory impairments over time. In conclusion, we have shown that brain tumor patients with cognitive complaints have high rates of episodic memory impairment. Retrieval deficits were detected in almost all patients presenting episodic memory impairments and were associated with storage deficits in about half the cases. Assessing brain tumor patients using an episodic memory sub-process test has high value and could be helpful in the design of rehabilitation. Supplementary Material Supplementary material is available at Archives of Clinical Neuropsychology online. Conflict of interest The authors have no conflicts of interest to disclose. References Armstrong, C. L., Stern, C. H., & Corn, B. W. ( 2001). Memory performance used to detect radiation effects on cognitive functioning. Applied Neuropsychology , 8, 129– 139. Google Scholar CrossRef Search ADS PubMed  Burger, L., Uittenhove, K., Lemaire, P., & Taconnat, L. ( 2017). Strategy difficulty effects in young and older adults’ episodic memory are modulated by inter-stimulus intervals and executive control processes. Acta Psychologica , 175, 50– 59. Google Scholar CrossRef Search ADS PubMed  Burgess, N., Maguire, E. A., & O’Keefe, J. ( 2002). The human hippocampus and spatial and episodic memory. Neuron , 35, 625– 641. Google Scholar CrossRef Search ADS PubMed  Buschke, H. ( 1984). Cued recall in amnesia. Journal of Clinical Neuropsychology , 6, 433– 440. Google Scholar CrossRef Search ADS PubMed  Douw, L., Klein, M., Fagel, S. S., van den Heuvel, J., Taphoorn, M. J. B., Aaronson, N. K., et al.  . ( 2009). Cognitive and radiological effects of radiotherapy in patients with low-grade glioma: Long-term follow-up. The Lancet Neurology , 8, 810– 818. Google Scholar CrossRef Search ADS PubMed  Duffau, H. ( 2014). Diffuse low-grade gliomas and neuroplasticity. Diagnostic and Interventional Imaging , 95, 945– 955. Google Scholar CrossRef Search ADS PubMed  Durand, T., Bernier, M. O., Léger, I., Taillia, H., Noël, G., Psimaras, D., et al.  . ( 2015). Cognitive outcome after radiotherapy in brain tumors. Current Opinion in Oncology , 27, 510– 515. Google Scholar CrossRef Search ADS PubMed  Giovagnoli, A. R., Meneses, R. F., Silvani, A., Molanesi, I., Fariselli, L., Salmaggi, A., et al.  . ( 2014). Quality of life and brain tumors: What beyond the clinical burden? Journal of Neurology , 261, 894– 904. Google Scholar CrossRef Search ADS PubMed  Grober, E., & Buschke, H. ( 1987). Genuine memory deficit in dementia. Developmental Psychology , 3, 13– 36. Jeong, W., Chung, C. K., & Kim, J. S. ( 2015). Episodic memory in aspects of large-scale brain networks. Frontiers in Human Neuroscience , 9, 454. Google Scholar CrossRef Search ADS PubMed  Klein, M. ( 2015). Treatment options and neurocognitive outcome in patients with diffuse low-grade glioma. Journal of Neurosurgical Science , 68, 389– 392. Lemos, R., Duro, D., Simões, M. R., & Santana, I. ( 2014). The free and cued selective reminding test distinguishes frontotemporal dementia from Alzheimer’s disease. Archive of Clinical Neuropsychology , 29, 670– 679. Google Scholar CrossRef Search ADS   Noll, K. R., Weinberg, J. S., Ziu, M., & Wefel, J. S. ( 2016). Verbal learning processes in patients with glioma of the left and right temporal lobes. Archive of Clinical Neuropsychology , 31, 37– 46. Google Scholar CrossRef Search ADS   Omuro, A. M., Ben-Porat, L. S., Panageas, K. S., Kim, A. K., Correa, D. D., Yahalom, J., et al.  . ( 2005). Delayed neurotoxicity in primary central nervous system lymphoma. Archive of Neurology , 62, 1595– 1600. Google Scholar CrossRef Search ADS   Tulving, E. ( 2004). Episodic memory: From mind to brain. Revue Neurologique (Paris) , 160, S9– S23. Google Scholar CrossRef Search ADS   Tulving, E., & Thomson, D. M. ( 1973). Encoding specificity and retrieval processes. Psychological Review , 80, 352– 373. Google Scholar CrossRef Search ADS   van der Linden, M., Coyette, F., Poitrenaud, J., Kalafat, M., Calicis, F., Wyns, C., et al.  . ( 2004). L’épreuve de rappel libre/rappel indicé à 16 item (RL/RI-16) in van der Linden M. In L’évaluation des troubles de la mémoire . Marseille: Edition Solal. van Loon, E. M. P., Heijenbrok-kal, M. H., van Loon, W. S., van den Bent, M. J., Vincent, A. J., de Koning, I., et al.  . ( 2015). Assessment methods and prevalence of cognitive dysfunction in patients with low-grade glioma: A systematic review. Journal of Rehabilitation Medicine , 47, 481– 488. Google Scholar CrossRef Search ADS PubMed  © The Author(s) 2018. Published by Oxford University Press. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Archives of Clinical Neuropsychology Oxford University Press

Loading next page...
 
/lp/ou_press/episodic-memory-impairments-in-primary-brain-tumor-patients-y0v14oYGx6
Publisher
Oxford University Press
Copyright
© The Author(s) 2018. Published by Oxford University Press. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com.
ISSN
0887-6177
eISSN
1873-5843
D.O.I.
10.1093/arclin/acx138
Publisher site
See Article on Publisher Site

Abstract

Abstract Objective Cognitive investigations in brain tumor patients have mostly explored episodic memory without differentiating between encoding, storage, and retrieval deficits. The aim of this study is to offer insight into the memory sub-processes affected in primary brain tumor patients and propose an appropriate assessment method. Method We retrospectively reviewed the clinical and memory assessments of 158 patients with primary brain tumors who had presented to our departments with cognitive complaints and were investigated using the Free and Cued Selective Reminding Test. Results Retrieval was the process of episodic memory most frequently affected, with deficits in this domain detected in 92% of patients with episodic memory impairments. Storage and encoding deficits were less prevalent, with impairments, respectively, detected in 41% and 23% of memory-impaired patients. The pattern of episodic memory impairment was similar across different tumor histologies and treatment modalities. Conclusion Although all processes of episodic memory were found to be impaired, retrieval was by far the most widely affected function. A thorough assessment of all three components of episodic memory should be part of the regular neuropsychological evaluation in patients with primary brain tumors. Memory decay, Brain neoplasms, Neurotoxicity, Cognitive assessment Introduction Cognitive impairments are a frequent and disabling complication in patients treated for brain tumors (Klein, 2015). Cognitive deficits can be related to tumor infiltration, tumor-related epilepsy, or to the effects of antitumoral treatments such as chemo- and radiotherapy. Medications such as antiepileptic drugs and steroids, which are frequently administered to these patients, have also been suggested to negatively influence cognitive function (Klein, 2015). Despite their dramatic impact on patients’ quality of life (Giovagnoli et al., 2014), the reported incidence of cognitive deficits in primary brain tumor patients varies widely across studies, from 20% to more than 80% (van Loon et al., 2015). Tests used and delays between diagnosis and assessment may partially explain these variations (Douw et al., 2009). Memory and attention have been consistently reported as the cognitive functions most frequently affected in brain tumor patients (Klein, 2015). Memory is a highly complex cognitive process that is normally split between different sub-process according to the nature of information being processed (e.g., episodic, semantic, procedural). Noll and colleagues recently showed that tumor location can specifically affect subtle cognitive processes and lead to distinct patterns of episodic memory dysfunction (Noll, Weinberg, Ziu, & Wefel, 2016). Episodic memory can be defined as the memory of personally experienced events within their temporo-spatial context: it allows for conscious recall of past experience of an event (what), its location (where), the time it occurred (when), and its emotional content (Tulving, 2004). Hermann Ebbinghaus (1850–1909) first introduced the concepts of encoding, storing/consolidation, and retrieval. These three sub-processes are the key phases in long-term memory formation and are essential for retrieving information when required. Encoding is the first phase of long-term memorization and consists of processing the characteristics of the information to be memorized in association with their environmental, cognitive, and emotional context. Storage refers to the consolidation of the encoded information in the memory system in order to prevent long-term forgetting. Retrieval is the final active phase in which the cognitive system transfers the information from long-term memory to working memory. The three sub-processes of episodic memory can take place alone or in combination, as has been highlighted in various pathological conditions (Lemos, Duro, Simões, & Santana, 2014). Therefore, an exhaustive investigation of episodic memory should include an assessment of all three of its components in order to detect which sub-processes are affected. So far, investigations in neuro-oncology have primarily explored episodic memory performance, but few studies have taken an interest in the mechanisms involved. The aim of this study is to offer insight into episodic memory sub-processes affected in primary brain tumor patients. Methods We performed a retrospective search of our institutional databases for patients with primary brain tumors referred to our departments (HIA de Percy, Clamart; GH Pitié-Salpêtrière, Paris, France) for suspected tumor- or treatment-related cognitive impairment during the period 2010–2015. Inclusion criteria were (1) a history of primary brain tumor with one of the following histologies: low-grade glioma (LGG), high-grade glioma (HGG), meningioma, or primary central nervous system lymphoma (PCNSL); (2) age ≥ 18 years old; and (3) episodic memory assessment performed using the French version of the Free and Cued Selective Reminding Test (van der Linden et al., 2004). Exclusion criteria were (1) the coexistence of other neurological or psychiatric disorders affecting cognitive function; and (2) personal history of a second neoplasm. Out of a total of 192 patients evaluated in our departments during this period, 158 matched the study criteria and could be included in this report. Clinical and neuropsychological assessments of the patients included in the study were collected and reviewed in detail. This research was carried out in accordance with the principles of the Declaration of Helsinki. Episodic Memory Assessment Verbal episodic memory was assessed using the Free and Cued Selective Reminding Test (FCSRT) (van der Linden et al. 2004). The FCSRT is widely used in francophone countries and has demonstrated its ability to detect and discriminate between encoding, storage, and retrieval impairments (Lemos et al., 2014). The test was adapted from the paradigm proposed by Buschke (Buschke, 1984) and Grober (Grober & Buschke, 1987) and is designed to manipulate encoding and retrieval conditions through the use of cues. It was developed on the basis of the observation that cues improve retrieval, especially when used not only during retrieval, but also during encoding (Tulving & Thomson, 1973). The FCSRT consists of several tasks. The first is a controlled encoding task during which the patient is asked to learn 16 words and to associate each with its semantic category (e.g., the word “judo” is associated with the category “sport”). It is followed by a free and cued recall task. During free recall, the patient is asked to give as many words as possible from the list of 16 words. During cued recall, the patient is provided with the category given at the encoding phase for each item not retrieved freely (e.g., what was the sport?). This is followed by a recognition task during which the patient is asked to identify the 16 original words, and reject distractors, from a list of 48 words. The test concludes with a delayed free and cued recall task 20 min after the end of the recognition task. The test is scored based on the number of items recalled freely compared to the total number of items retrieved (free and cued recall). Between each recall task, the patient must perform a distraction task, consisting of counting for 20 s. Patients’ scores in our study were compared with normative data from 483 healthy controls (Van der Linder et al., 2004) matched for age, gender, and education. Each of the nine scores recorded (number of word recalled at immediate recall), free recalls (1, 2, 3, delayed) and cued recalls (1, 2, 3, delayed) were considered abnormal when it corresponded to a performance equal to or under the fifth percentile of the healthy controls normative data (van der Linden et al., 2004). An encoding deficit was diagnosed when the immediate recall was abnormal (the assumption being that the items were not present in the working memory immediately after they have been red, and so not encoded). A failure in free recalls corresponded to at least 2/4 abnormal scores and a failure in cued recalls corresponded to at least 2/4 abnormal scores (the test is composed of four free and four cued recalls). A storage deficit was diagnosed in the case of a failure in free recalls associated with a failure in cued recalls. This means that the cues didn’t improve the number of items recalled, assuming that the items was not stored. A retrieval deficit was diagnosed in the case of a failure in free recalls isolated (normal cued recalls). Indeed, the cues improved the number of items recalled comparably to healthy controls, giving a proof that items was stored in the memory but not available at the moment. Furthermore, an association of storage and retrieval deficit was diagnosed in the case of a failure in free recalls and a failure in cued recalls, but with limited improvement (incomparably to healthy controls) of the total number of items recalled with cues. In other words, items were retrieved with difficulties (because cues helped after a failure in free recalls), but was also not stored as expected (because cues did not help to reach normal number of recalls, meaning that items was not entirely stored). To help in the understanding of the FCRST design, Supplementary Fig. 1 gives an overview of the protocol, the criteria and the interpretation of the different pattern of results. Statistical Analysis Descriptive analyses were used for FCSRT scores and for patient and tumor characteristics. Patients with episodic memory impairments were compared to non-impaired patients to search for differences in demographic or tumor characteristics. The t-test was used for continuous variables. The χ2 or Fisher’s exact test was used for categorical comparison. Univariate and multiple logistic regressions were used to explore the association between the presence of episodic memory deficits and a number of patient and tumor characteristics. All data analyses were performed using R statistical software. Results Clinical characteristics of our 158 patients are summarized in Table 1. Median age at the time of neuropsychological assessment was 58 (range 22–87). Tumor histology was as follows: HGG in 79 patients (50%), LGG in 30 patients (19%), meningioma in 33 patients (21%), and PCNSL in 16 patients (10%). All patients had supratentorial neoplasms, the frontal lobes being the most frequent location. Tumor treatment included surgery, radiotherapy, and/or chemotherapy using various schemes and combinations, based on tumor histology and the patient’s clinical conditions. Overall, 78 patients received surgical resection (50%), 64 patients were irradiated (40%), and another 64 received chemotherapy (40%). Table 1. Patients’ and tumor characteristics Demographics  Whole group (n = 158)  Non-impaired group (n = 92)  Impaired group (n = 66)  p-valuea  Male/female  86 (54%)/72 (46%)  50 (54%)/42 (46%)  36 (54%)/30 (46%)  1  Age (median, range)  58 (22–87)  55 (26–87)  62 (22–87)  .082  Education   <12 years  40 (25%)  26 (28%)  14 (21%)  .556   12 years  20 (13%)  12 (13%)  8 (12%)     >12 years  98 (62%)  54 (59%)  44 (67%)    Handedness   Left  5 (3%)  2 (2%)  3 (5%)  .045   Ambidextrous  6 (4%)  6 (6%)  0     Right  147 (93%)  84 (92%)  63 (95%)    Tumor characteristics   Histology    LGG  30 (19%)  9 (10%)  21 (32%)  .001    HGG  79 (50%)  48 (52%)  31 (47%)      Meningioma  33 (21%)  27 (29%)  6 (9%)      PCNSL  16 (10%)  8 (9%)  8 (12%)    Locationb   Left frontal  43 (27%)  21 (23%)  22 (33%)  .04   Right frontal  99 (62%)  70 (76%)  29 (44%)     Left temporal  18 (11%)  9 (10%)  9 (14%)     Right temporal  35 (22%)  25 (27%)  10 (15%)     Left parietal  12 (8%)  6 (6%)  6 (9%)     Right parietal  16 (10%)  8 (9%)  8 (12%)     Left occipital  4 (2%)  3 (3%)  1 (1%)     Right occipital  13 (8%)  11 (12%)  2 (2%)     Corpus Callosum  15 (9%)  6 (6%)  9 (14%)     Basal ganglia  16 (10%)  8 (9%)  8 (12%)    Prior antitumoral treatmentsc   Biopsy alone  40 (25%)  21 (23%)  19 (29%)  .013   Exeresis alone  41 (26%)  33 (36%)  8 (12%)     RT alone  6 (4%)  4 (4%)  2 (3%)     CT alone  11 (7%)  8 (9%)  3 (4%)     Exeresis + CT  2 (1%)  0  2 (3%)     Exeresis + RT  7 (4%)  4 (4%)  3 (4%)     RT + CT  23 (14%)  10 (11%)  13 (20%)     Exeresis + RT + CT  28 (18%)  12 (13%)  16 (24%)     Delay since RT (mean, SD)  7.9 years (±6.6)  7.7 years (±6.2)  8.2 years (±8.3)  .835   Delay since RT (range)  0.5–20 years  0.5–12 years  0.5–20 years    Concomitant medications   Antiepileptic drugs  91 (58%)  51 (55%)  40 (61%)  .627   Steroids  35 (22%)  22 (24%)  13 (20%)  .663  Demographics  Whole group (n = 158)  Non-impaired group (n = 92)  Impaired group (n = 66)  p-valuea  Male/female  86 (54%)/72 (46%)  50 (54%)/42 (46%)  36 (54%)/30 (46%)  1  Age (median, range)  58 (22–87)  55 (26–87)  62 (22–87)  .082  Education   <12 years  40 (25%)  26 (28%)  14 (21%)  .556   12 years  20 (13%)  12 (13%)  8 (12%)     >12 years  98 (62%)  54 (59%)  44 (67%)    Handedness   Left  5 (3%)  2 (2%)  3 (5%)  .045   Ambidextrous  6 (4%)  6 (6%)  0     Right  147 (93%)  84 (92%)  63 (95%)    Tumor characteristics   Histology    LGG  30 (19%)  9 (10%)  21 (32%)  .001    HGG  79 (50%)  48 (52%)  31 (47%)      Meningioma  33 (21%)  27 (29%)  6 (9%)      PCNSL  16 (10%)  8 (9%)  8 (12%)    Locationb   Left frontal  43 (27%)  21 (23%)  22 (33%)  .04   Right frontal  99 (62%)  70 (76%)  29 (44%)     Left temporal  18 (11%)  9 (10%)  9 (14%)     Right temporal  35 (22%)  25 (27%)  10 (15%)     Left parietal  12 (8%)  6 (6%)  6 (9%)     Right parietal  16 (10%)  8 (9%)  8 (12%)     Left occipital  4 (2%)  3 (3%)  1 (1%)     Right occipital  13 (8%)  11 (12%)  2 (2%)     Corpus Callosum  15 (9%)  6 (6%)  9 (14%)     Basal ganglia  16 (10%)  8 (9%)  8 (12%)    Prior antitumoral treatmentsc   Biopsy alone  40 (25%)  21 (23%)  19 (29%)  .013   Exeresis alone  41 (26%)  33 (36%)  8 (12%)     RT alone  6 (4%)  4 (4%)  2 (3%)     CT alone  11 (7%)  8 (9%)  3 (4%)     Exeresis + CT  2 (1%)  0  2 (3%)     Exeresis + RT  7 (4%)  4 (4%)  3 (4%)     RT + CT  23 (14%)  10 (11%)  13 (20%)     Exeresis + RT + CT  28 (18%)  12 (13%)  16 (24%)     Delay since RT (mean, SD)  7.9 years (±6.6)  7.7 years (±6.2)  8.2 years (±8.3)  .835   Delay since RT (range)  0.5–20 years  0.5–12 years  0.5–20 years    Concomitant medications   Antiepileptic drugs  91 (58%)  51 (55%)  40 (61%)  .627   Steroids  35 (22%)  22 (24%)  13 (20%)  .663  LGG = low-grade glioma; HGG = high-grade glioma; PCNSL = primary central nervous system lymphoma; RT = radiotherapy; CT = chemotherapy. at-Tests, Pearson χ2 or Fisher’s exact test for comparisons between non-impaired and impaired group. bAll patients had a unique tumor but 51 patients had a tumor involving several brain lobes. cRadiotherapy, whenever performed, was done using a 3D-conformational technique. Table 1. Patients’ and tumor characteristics Demographics  Whole group (n = 158)  Non-impaired group (n = 92)  Impaired group (n = 66)  p-valuea  Male/female  86 (54%)/72 (46%)  50 (54%)/42 (46%)  36 (54%)/30 (46%)  1  Age (median, range)  58 (22–87)  55 (26–87)  62 (22–87)  .082  Education   <12 years  40 (25%)  26 (28%)  14 (21%)  .556   12 years  20 (13%)  12 (13%)  8 (12%)     >12 years  98 (62%)  54 (59%)  44 (67%)    Handedness   Left  5 (3%)  2 (2%)  3 (5%)  .045   Ambidextrous  6 (4%)  6 (6%)  0     Right  147 (93%)  84 (92%)  63 (95%)    Tumor characteristics   Histology    LGG  30 (19%)  9 (10%)  21 (32%)  .001    HGG  79 (50%)  48 (52%)  31 (47%)      Meningioma  33 (21%)  27 (29%)  6 (9%)      PCNSL  16 (10%)  8 (9%)  8 (12%)    Locationb   Left frontal  43 (27%)  21 (23%)  22 (33%)  .04   Right frontal  99 (62%)  70 (76%)  29 (44%)     Left temporal  18 (11%)  9 (10%)  9 (14%)     Right temporal  35 (22%)  25 (27%)  10 (15%)     Left parietal  12 (8%)  6 (6%)  6 (9%)     Right parietal  16 (10%)  8 (9%)  8 (12%)     Left occipital  4 (2%)  3 (3%)  1 (1%)     Right occipital  13 (8%)  11 (12%)  2 (2%)     Corpus Callosum  15 (9%)  6 (6%)  9 (14%)     Basal ganglia  16 (10%)  8 (9%)  8 (12%)    Prior antitumoral treatmentsc   Biopsy alone  40 (25%)  21 (23%)  19 (29%)  .013   Exeresis alone  41 (26%)  33 (36%)  8 (12%)     RT alone  6 (4%)  4 (4%)  2 (3%)     CT alone  11 (7%)  8 (9%)  3 (4%)     Exeresis + CT  2 (1%)  0  2 (3%)     Exeresis + RT  7 (4%)  4 (4%)  3 (4%)     RT + CT  23 (14%)  10 (11%)  13 (20%)     Exeresis + RT + CT  28 (18%)  12 (13%)  16 (24%)     Delay since RT (mean, SD)  7.9 years (±6.6)  7.7 years (±6.2)  8.2 years (±8.3)  .835   Delay since RT (range)  0.5–20 years  0.5–12 years  0.5–20 years    Concomitant medications   Antiepileptic drugs  91 (58%)  51 (55%)  40 (61%)  .627   Steroids  35 (22%)  22 (24%)  13 (20%)  .663  Demographics  Whole group (n = 158)  Non-impaired group (n = 92)  Impaired group (n = 66)  p-valuea  Male/female  86 (54%)/72 (46%)  50 (54%)/42 (46%)  36 (54%)/30 (46%)  1  Age (median, range)  58 (22–87)  55 (26–87)  62 (22–87)  .082  Education   <12 years  40 (25%)  26 (28%)  14 (21%)  .556   12 years  20 (13%)  12 (13%)  8 (12%)     >12 years  98 (62%)  54 (59%)  44 (67%)    Handedness   Left  5 (3%)  2 (2%)  3 (5%)  .045   Ambidextrous  6 (4%)  6 (6%)  0     Right  147 (93%)  84 (92%)  63 (95%)    Tumor characteristics   Histology    LGG  30 (19%)  9 (10%)  21 (32%)  .001    HGG  79 (50%)  48 (52%)  31 (47%)      Meningioma  33 (21%)  27 (29%)  6 (9%)      PCNSL  16 (10%)  8 (9%)  8 (12%)    Locationb   Left frontal  43 (27%)  21 (23%)  22 (33%)  .04   Right frontal  99 (62%)  70 (76%)  29 (44%)     Left temporal  18 (11%)  9 (10%)  9 (14%)     Right temporal  35 (22%)  25 (27%)  10 (15%)     Left parietal  12 (8%)  6 (6%)  6 (9%)     Right parietal  16 (10%)  8 (9%)  8 (12%)     Left occipital  4 (2%)  3 (3%)  1 (1%)     Right occipital  13 (8%)  11 (12%)  2 (2%)     Corpus Callosum  15 (9%)  6 (6%)  9 (14%)     Basal ganglia  16 (10%)  8 (9%)  8 (12%)    Prior antitumoral treatmentsc   Biopsy alone  40 (25%)  21 (23%)  19 (29%)  .013   Exeresis alone  41 (26%)  33 (36%)  8 (12%)     RT alone  6 (4%)  4 (4%)  2 (3%)     CT alone  11 (7%)  8 (9%)  3 (4%)     Exeresis + CT  2 (1%)  0  2 (3%)     Exeresis + RT  7 (4%)  4 (4%)  3 (4%)     RT + CT  23 (14%)  10 (11%)  13 (20%)     Exeresis + RT + CT  28 (18%)  12 (13%)  16 (24%)     Delay since RT (mean, SD)  7.9 years (±6.6)  7.7 years (±6.2)  8.2 years (±8.3)  .835   Delay since RT (range)  0.5–20 years  0.5–12 years  0.5–20 years    Concomitant medications   Antiepileptic drugs  91 (58%)  51 (55%)  40 (61%)  .627   Steroids  35 (22%)  22 (24%)  13 (20%)  .663  LGG = low-grade glioma; HGG = high-grade glioma; PCNSL = primary central nervous system lymphoma; RT = radiotherapy; CT = chemotherapy. at-Tests, Pearson χ2 or Fisher’s exact test for comparisons between non-impaired and impaired group. bAll patients had a unique tumor but 51 patients had a tumor involving several brain lobes. cRadiotherapy, whenever performed, was done using a 3D-conformational technique. Patients were evaluated at variable intervals after the tumor diagnosis (range 6–240 months, median 92). At the time of cognitive assessment, 58% of patients were taking antiepileptic drugs and 22% were taking steroids; none of the patients was actively receiving antitumoral treatments as their tumors were considered in remission. Overall, 66 (42%) of the patients in our study series were found to have episodic memory impairments. Retrieval deficits were by far the most frequent impairment, affecting 39% of patients in the whole cohort and 92% of patients with episodic memory impairments (isolated or in association with encoding and/or storage deficits). Encoding and storage deficits were less prevalent, respectively, affecting 9% and 17% of patients in the whole cohort, and 23% and 41% of patients with episodic memory impairments (Fig. 1A). Encoding deficits were associated with both storage (Logistic regression: OR 5.38, β = 1.68, 95% IC 2.07–13.85, p = .003) and retrieval (Logistic regression: OR 3.6, β = 1.28, 95% IC 1.44–9.82, p = .02) deficits. Fig. 1B shows the frequencies of encoding, storage, and retrieval impairments across different tumor histologies. Fig. 1. View largeDownload slide Frequency of encoding, storage, and retrieval deficits. (A) Frequency of encoding, storage, and retrieval impairments in the whole cohort (light gray) and in the impaired patient group (dark gray). (B) frequency of encoding (light gray), storage (medium gray), and retrieval (dark gray) impairments based on tumor histology (whole cohort). Fig. 1. View largeDownload slide Frequency of encoding, storage, and retrieval deficits. (A) Frequency of encoding, storage, and retrieval impairments in the whole cohort (light gray) and in the impaired patient group (dark gray). (B) frequency of encoding (light gray), storage (medium gray), and retrieval (dark gray) impairments based on tumor histology (whole cohort). We then compared patients with episodic memory impairments (impaired patients) to patients without any episodic memory impairments (non-impaired patients) (Table 1). The group of impaired patients had a higher prevalence of LGG and a lower prevalence of meningioma histology (χ2: p = .001). Impaired patients were also more likely to have received radio-chemotherapy (concomitantly ± sequentially, or only sequentially) and less likely to have undergone exeresis alone (χ2: p = .013). Logistic regression confirmed that a histology of LGG (Logistic regression: OR 4.304, β = 1.459, 95% IC = 1.86–10.63, p = .0009) and prior radio-chemotherapy treatment (Logistic regression: OR 2.494, β = 0.914, 95% IC 1.26–4.98, p = .008) were factors significantly associated with memory impairment, after adjusting for other patient and tumor characteristics. No correlations between specific tumor locations and types of episodic memory impairment were found, except for the association of encoding deficits with corpus callosum infiltration (Logistic regression: OR 4.36, β = 1.68, 95% IC 1.37–12.58, p = .02). However, when we pooled results for patients with tumors located in the left hemisphere and compared them with pooled results for patients with tumors located in the right hemisphere, we found that memory impairments appeared to be significantly more prevalent in patients with tumors located in the left hemisphere (χ2: 48% vs. 31%, p = .013). No association between the frequency or nature of episodic memory impairments and clinical characteristics such as age, gender, education, or concomitant medications was detected in our study. Discussion In the present study, we explored all three components of episodic memory in a large series of patients with primary brain tumors. Episodic memory deficits were highly prevalent in our cohort (42%), suggesting that this domain should be systematically investigated in these patients, as has already been recommended (Klein, 2015). To our knowledge, we are the first to report that retrieval was the component of episodic memory most frequently affected, with deficits in this process found in 92% of patients with episodic memory impairments (isolated or in association with encoding and/or storage deficit). Storage deficits were also present in a non-negligible proportion of patients (41% of patients with impairments). The extremely high prevalence of retrieval deficits detected in impaired patients is consistent with what we observe on a daily basis in the neuro-oncology department. Although the reasons for this finding are still unknown, we can speculate that retrieval is more frequently affected because it recruits a larger neural network (Jeong, Chung, & Kim, 2015), which is more likely to be disrupted as a result of tumor infiltration or antitumoral treatments (surgery, radiotherapy, and possibly chemotherapy). Furthermore, the frequency of impairments reported here is probably overestimated because of a recruitment bias in favor of patients presenting with cognitive complaints. These results should not be considered as thorough epidemiological data, but rather as a first overview of impairment patterns. Episodic memory can also be influenced by other cognitive processes such as selective attention, working memory, and organizational skills (Burger, Uittenhove, Lemaire, & Taconnat, 2017). One study conducted on a small group of patients with LGG addressed this particular issue, concluding that the episodic memory impairments observed in the cohort could not be attributed to deficits in these other domains (Armstrong, Stern, & Corn, 2001). Still, validation of these results in a larger prospective setting is needed. Episodic memory is a highly complex cognitive process that involves several areas and networks. It has recently been postulated that, along with the medio-temporal/hippocampal region, other areas such as the prefrontal and the parietal cortex may play an important role with regard to encoding and retrieval (Jeong et al., 2015). In our cohort, we did not observe a relation between any one single tumor location and type of episodic memory impairment, except for patients with corpus callosum infiltration who did worse at encoding, probably due to attention issues. However, patients with tumors in the left hemisphere had a significantly higher prevalence of episodic memory impairments compared to patients with tumors in the right hemisphere (χ2: p = .013). These results could be explained by the nature of the material used in the FCRST. Right structures may be more involved in spatial memory, which is not tested through the FCRST, while left structures could be more involved in context-dependent and autobiographic memory (Burgess, Maguire, & O’Keefe, 2002). Our study included patients with primary brain tumors corresponding to different histologies. We found a higher prevalence of storage and retrieval deficits in LGG patients compared to patients with other tumor types. Although tumor infiltrative behavior is an important factor to consider due to its capacity to disrupt cortical–subcortical connectivity (Duffau, 2014), the high prevalence of storage and retrieval deficits in patients with LGG observed in our series is more likely attributable to a selection bias. LGG patients live longer and, in most cases, were referred to our center for cognitive complaints after treatment (Douw et al., 2009). Still, despite these differences in prevalence, the profile of episodic memory impairments was similar across tumor types, suggesting there could be a common pathogenetic basis that transcends histology. Patients who received radio-chemotherapy appeared more prone to developing memory impairments than patients who did not. Brain irradiation remains a recognized risk factor for late neurotoxicity (Klein, 2015) due to its capacity to reduce neuronal plasticity and induce chronic progressive leukoencephalopathy (Omuro et al., 2005). Furthermore, the concomitant administration of radiation and chemotherapy could have a synergistic neurotoxic effect (Durand et al., 2015). Because of the cross-sectional design of our study, and given that most of our patients were assessed after receiving adjuvant treatments, it is impossible for us to discriminate between the effect of the tumor itself and the effect of antitumoral treatments. Further prospective studies are needed to investigate the possible influence of other cognitive functions such as attention and organizational skills (Noll et al., 2016) on memory deficits and the evolution of episodic memory impairments over time. In conclusion, we have shown that brain tumor patients with cognitive complaints have high rates of episodic memory impairment. Retrieval deficits were detected in almost all patients presenting episodic memory impairments and were associated with storage deficits in about half the cases. Assessing brain tumor patients using an episodic memory sub-process test has high value and could be helpful in the design of rehabilitation. Supplementary Material Supplementary material is available at Archives of Clinical Neuropsychology online. Conflict of interest The authors have no conflicts of interest to disclose. References Armstrong, C. L., Stern, C. H., & Corn, B. W. ( 2001). Memory performance used to detect radiation effects on cognitive functioning. Applied Neuropsychology , 8, 129– 139. Google Scholar CrossRef Search ADS PubMed  Burger, L., Uittenhove, K., Lemaire, P., & Taconnat, L. ( 2017). Strategy difficulty effects in young and older adults’ episodic memory are modulated by inter-stimulus intervals and executive control processes. Acta Psychologica , 175, 50– 59. Google Scholar CrossRef Search ADS PubMed  Burgess, N., Maguire, E. A., & O’Keefe, J. ( 2002). The human hippocampus and spatial and episodic memory. Neuron , 35, 625– 641. Google Scholar CrossRef Search ADS PubMed  Buschke, H. ( 1984). Cued recall in amnesia. Journal of Clinical Neuropsychology , 6, 433– 440. Google Scholar CrossRef Search ADS PubMed  Douw, L., Klein, M., Fagel, S. S., van den Heuvel, J., Taphoorn, M. J. B., Aaronson, N. K., et al.  . ( 2009). Cognitive and radiological effects of radiotherapy in patients with low-grade glioma: Long-term follow-up. The Lancet Neurology , 8, 810– 818. Google Scholar CrossRef Search ADS PubMed  Duffau, H. ( 2014). Diffuse low-grade gliomas and neuroplasticity. Diagnostic and Interventional Imaging , 95, 945– 955. Google Scholar CrossRef Search ADS PubMed  Durand, T., Bernier, M. O., Léger, I., Taillia, H., Noël, G., Psimaras, D., et al.  . ( 2015). Cognitive outcome after radiotherapy in brain tumors. Current Opinion in Oncology , 27, 510– 515. Google Scholar CrossRef Search ADS PubMed  Giovagnoli, A. R., Meneses, R. F., Silvani, A., Molanesi, I., Fariselli, L., Salmaggi, A., et al.  . ( 2014). Quality of life and brain tumors: What beyond the clinical burden? Journal of Neurology , 261, 894– 904. Google Scholar CrossRef Search ADS PubMed  Grober, E., & Buschke, H. ( 1987). Genuine memory deficit in dementia. Developmental Psychology , 3, 13– 36. Jeong, W., Chung, C. K., & Kim, J. S. ( 2015). Episodic memory in aspects of large-scale brain networks. Frontiers in Human Neuroscience , 9, 454. Google Scholar CrossRef Search ADS PubMed  Klein, M. ( 2015). Treatment options and neurocognitive outcome in patients with diffuse low-grade glioma. Journal of Neurosurgical Science , 68, 389– 392. Lemos, R., Duro, D., Simões, M. R., & Santana, I. ( 2014). The free and cued selective reminding test distinguishes frontotemporal dementia from Alzheimer’s disease. Archive of Clinical Neuropsychology , 29, 670– 679. Google Scholar CrossRef Search ADS   Noll, K. R., Weinberg, J. S., Ziu, M., & Wefel, J. S. ( 2016). Verbal learning processes in patients with glioma of the left and right temporal lobes. Archive of Clinical Neuropsychology , 31, 37– 46. Google Scholar CrossRef Search ADS   Omuro, A. M., Ben-Porat, L. S., Panageas, K. S., Kim, A. K., Correa, D. D., Yahalom, J., et al.  . ( 2005). Delayed neurotoxicity in primary central nervous system lymphoma. Archive of Neurology , 62, 1595– 1600. Google Scholar CrossRef Search ADS   Tulving, E. ( 2004). Episodic memory: From mind to brain. Revue Neurologique (Paris) , 160, S9– S23. Google Scholar CrossRef Search ADS   Tulving, E., & Thomson, D. M. ( 1973). Encoding specificity and retrieval processes. Psychological Review , 80, 352– 373. Google Scholar CrossRef Search ADS   van der Linden, M., Coyette, F., Poitrenaud, J., Kalafat, M., Calicis, F., Wyns, C., et al.  . ( 2004). L’épreuve de rappel libre/rappel indicé à 16 item (RL/RI-16) in van der Linden M. In L’évaluation des troubles de la mémoire . Marseille: Edition Solal. van Loon, E. M. P., Heijenbrok-kal, M. H., van Loon, W. S., van den Bent, M. J., Vincent, A. J., de Koning, I., et al.  . ( 2015). Assessment methods and prevalence of cognitive dysfunction in patients with low-grade glioma: A systematic review. Journal of Rehabilitation Medicine , 47, 481– 488. Google Scholar CrossRef Search ADS PubMed  © The Author(s) 2018. Published by Oxford University Press. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com.

Journal

Archives of Clinical NeuropsychologyOxford University Press

Published: Jan 4, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off